Photophysical, dynamic and redox behavior of tris(2,6-diisopropylphenyl) phosphine

View Article Online / Journal Homepage / Table of Contents for this issue

PAPER

www.rsc.org/njc | New Journal of Chemistry

Photophysical, dynamic and redox behavior of

tris(2,6-diisopropylphenyl)phosphinew

a

b

a

b

´

Rene T. Boere,* Alan M. Bond,* Steve Cronin, Noel W. Duﬀy,

Paul Hazendonk,^aJason D. Masuda,^aKyle Pollard,^aTracey L. Roemmele,^a

Peter Tran^aand Yuankui Zhang^a

Received (in Montpellier, France) 27th June 2007, Accepted 14th September 2007

First published as an Advance Article on the web 4th October 2007

DOI: 10.1039/b709602j

The title phosphine, Dipp₃P, was synthesized using an aryl copper reagent and the structure

determined by X-ray crystallography (R = 2.94%): d(P–C) = 1.852(1) A, +C–P–C =

111.88(5)1. In hexane solution, the electronic spectrum displays 3 bands [326 (9.3), 254 (8.7), 205

(11.4) nm (log|e|)] and the ﬂuorescence spectrum has a Stokes shift of 129 kJ mol^ꢀ1. NMR: (d)

1

³¹P = ꢀ49.7 ppm in solution and ꢀ49.5 in the solid (CP-MAS). Room temperature H and ¹³C

spectra reﬂect D₃symmetry, changing below ꢀ30 1C to C₃. A variable temperature NMR study

provided an activation enthalpy of 49(ꢁ1) kJ mol^ꢀ1and entropy of 24–27(ꢁ5) J mol^ꢀ1

K

^ꢀ1. An

energy surface calculation using HF/3-21G theory discovered a single low-energy path describing

pyramidal inversion through a transition state that is close to D₃geometry. The B3LYP/6-31G(d)

calculated barrier to planarization is 37.5 kJ mol^ꢀ1. Voltammetric studies employing cyclic,

rotating disk, steady state and Fourier Transform ac methods conﬁrm a fully chemically

reversible one-electron oxidation of Dipp₃P to Dipp₃P⁺at +0.18 (CH₃CN–ⁿBu₄NPF₆) and

ꢂ

+0.09 (CH₂Cl₂–ⁿBu₄NPF₆) V vs. Fc^+/0(Fc = ferrocene). The diﬀusion coeﬃcient for Dipp₃P is

1.0–1.2 ꢃ 10⁵cm²s^ꢀ1. The electrode process displays quasi-reversible electron transfer kinetics

[k_sE 0.01 (CH₂Cl₂) to 0.08 (_ꢂCH₃CN) cm s^ꢀ1]. Optically transparent thin layer electrolysis

reversibly generates Dipp₃P⁺in CH₂Cl₂–ⁿBu₄NPF₆[UV-Vis: 498 (3.31), 456 (3.29), 373 (4.04),

357 (3.84), 341 (3.49), 296 (3.78), 385 (3.91), 251 (3.99) nm (log|e|)]. The EPR spectrum of

Dipp₃P⁺in solution is a doublet (a(P) = 23.9 mT, g = 2.008), and in frozen solution is axial

ꢂ

(a_J= 42.6 mT, g_J= 2.0045; a_>= 12.7 mT, g_>= 2.0085 mT).

both intrinsically and especially in a wide variety of catalytic

Introduction

applications,⁶and recent advances have been reviewed.^6aThe

study of phosphorus redox centers has also received consider-

able attention in recent years,⁷particularly when linked che-

mically to ferrocene units.⁸Bulky phosphines are also well-

known to have restricted intra-molecular dynamics, an old

topic which has received renewed attention.⁹

Trialkyl- and triarylphosphines¹are extremely important

ligands in organometallic coordination chemistry and many

of their characteristics have been systematically categorized by

steric bulk ever since the introduction of Tolman’s cone angle

classiﬁcation.²For many years, the record cone angle of 2121

was held by trimesitylphosphine (Mes₃P), 1.^3,4Despite its

bulkiness, Mes₃P can still form metal complexes, albeit with

low coordination numbers, such as in linear Mes₃PAuCl.⁵

There is a great deal of current interest in bulky phosphines,

We have been interested for some time in the thermody-

namic, kinetic and geometric consequences of high steric bulk.

For example, we have designed several ligand systems in which

bulky substituents are employed at low-coordinate nitrogen

and phosphorus centers.¹⁰In the course of our work we

prepared the new bulky primary phosphine 2,6-diisopropyl-

phenylphosphine (DippPH₂) (Dipp = 2,6-diisopropylphenyl;

for a full list of abbreviations, see ref. 4).¹¹This led us to

consider the intriguing possibility as to whether the even more

bulky tris Dipp-substituted phosphorus compound could be

prepared, and if so what properties it would display.

^aUniversity of Lethbridge, Department of Chemistry and

Biochemistry, The University of Lethbridge, Lethbridge, AB Canada

T1K 3M4. E-mail: boere@uleth.ca; Fax: +1-403-329-2057;

Tel: +1-403-329-2045

^bSchool of Chemistry, Monash University, Clayton, Victoria 3800,

Australia. E-mail: Alan.Bond@sci.monash.edu.au; Fax: +61 3 9905

4597; Tel: +61 3 9905 1338

w Electronic supplementary information (ESI) available: Table S1:

Full geometric details. Table S2: Electronic spectra of Dipp₃P and

During the course of our work, the closely-related tris(2,4,6-

triisopropylphenyl)phosphine (Tripp₃P) was reported by Sa-

saki, Yoshifuji and co-workers,¹²and several interesting mixed

Dipp/Tripp substituent systems in which one ring is substi-

tuted at the para site have been prepared in order to investigate

possible redox interactions.^7f–mDipp₃P and Tripp₃P are the

most sterically congested triaryl phosphines reported to date.¹³

Dipp₃P⁺and Ph₃P analogs. Table S3: Calculated hfc constants.

ꢂ

Table S4: CV Data. Table S5: Diﬀusion coeﬃcient values for Dipp₃P.

Tables S6, S7: Concentration dependence of E_f1. Fig. S1: Static SS ³¹

P

NMR spectrum. Fig. S2: Solution ¹H NMR of Dipp₃P. Figs. S3–5:

Eyring plots for DNMR. Fig. S6: Simulated and experimental RDE

voltammograms. Fig. S7: Spin density plot of the radical cation. Fig.

S8: Walsh diagram for planar and pyramidal EH₃. See DOI: 10.1039/

b709602j

ꢄc

This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientiﬁque 2008

214 | New J. Chem., 2008, 32, 214–231

View Article Online

and Yoshifuji,¹²though persistent radicals from sterically

+

congested species such as Mes₃P⁺^ꢂ, Duryl₃P and Xyl₃P⁺

ꢂ

have been claimed for some considerable time.²⁷Last but not

least, in order to overcome considerable confusion in the

literature, we have measured potentials for the 0/+1 process

of several bulky triaryl phosphines under identical conditions

in order to put their potentials on a common scale.

Fig. 1 (a) In a 2-substituted triarylphosphine, the endo oriented

hydrogen atoms exert mild steric pressure, but the exo substituents

(Ex) exert reverse steric pressure compressing +C–P–C. (b) In a 2,6-

disubstituted triarylphosphine, severe steric pressure from the endo

substituents (En) causes an increase in +C–P–C.¹²

Results

Synthesis and properties of Dipp₃P

Dipp₃P can be produced in reasonable yield (60–70% of

crystalline material) by converting the Grignard reagent pre-

pared from DippBr into (DippCu)_x,²⁸and reacting in situ with

PCl₃.²⁹There are very few reports on the use of organocopper

reagents in the synthesis of triorganophosphines before our

work and that of Sasaki and Yoshifuji.¹²Aranyos et al. used a

biphenylcopper reagent to prepare 2-(di-tert-butylphosphino)-

biphenyl.³⁰Greenﬁeld and Gilbertson used a complex se-

quence of organozinc followed by lithium cuprate reagents

to prepare primary and secondary amino acid derivatives of

Ph₂PX.³¹Langer et al. have shown the general utility of using

organozinc iodides in reactions with Ph₂PCl.³²

Thus, a detailed study of the intrinsic properties of these

systems is of particular current relevance.^6aIn a recent con-

ference proceeding, we introduced a distinction between steric

pressure and steric shielding in triaryl phosphines.¹⁴Thus, a

single set of ortho substituents (2-position) can provide con-

siderable steric shielding for the phosphorus lone pair, and

consequently a large Tolman cone angle, without any ﬂatten-

ing of the phosphorus pyramid (Fig. 1(a)). But the presence of

substituents in both the 2- and the 6-position of the aromatic

ring induces steric pressure which leads to actual ﬂattening of

the phosphorus pyramid (Fig. 1(b)).¹⁴For example, in

(2-ⁱPrC₆H₄)₃P, for which the preferred conformation has all

We were unable to measure a melting point for Dipp₃P as it

sublimes out of the (capillary) melting point apparatus at

atmospheric pressure, reﬂecting both its high thermal stability

P

the substituents in the exo position, {+CPC} = 306.51, a

value which is actually less than in Ph₃P: 308.11.^15,16But in

Dipp₃P and Tripp₃P, the steric pressure generated by the three

and considerable volatility. The radical cation Dipp₃P⁺

ꢂ

P

observed in the electron impact mass spectrum also appears

to be very stable: the parent molecular ion is not only the base

peak in the 70 eV spectrum but is virtually the only ion

recorded. The next most abundant fragment (from loss of a

single isopropyl group) has a relative intensity of 9%, and two

bulky endo substituents causes an increase in {+CPC} to

335.6 and 334.41, respectively. A synergistic eﬀect operates

such that the ﬂattening caused by the endo substituents causes

the exo substituents to be pushed closer around the phos-

phorus atom and provides enhanced shielding for the lone pair

in these compounds with the same exo substituents.¹⁴

+

others assigned to C₆H₁₀P⁺and C₇H₇are each at 4%.

Grutzmacher and Kirchhoﬀ have discussed the mass spectra

¨

We now report full details of the synthesis, chemical proper-

ties, structure as derived from X-ray diﬀraction and computa-

tion, photophysics, solution dynamics and redox behavior of

Dipp₃P. We demonstrate that it has numerous exceptional

properties, which are largely attributable to either steric

pressure or steric shielding, but which also involve the induc-

tive inﬂuence of the bulky alkyl substituents. The electroche-

mical investigation employs inter alia the new large amplitude

form of FT ⁴ac-voltammetry which is under development in

one of our laboratories.¹⁷

of triphenylpnicogens in terms of the element–carbon bond

strengths.³³By way of comparison, in the EI mass spectrum of

triphenylphosphine, signiﬁcant fragments are found for

Ph₂P⁺(67%) and PhP⁺(33%) and many other fragments

are also produced.³³

Dipp₃P is remarkably unreactive. It does not appear to react

with O₂from air, even at elevated temperature in solution, nor

can the oxide be prepared with H₂O₂in boiling acetone.

Reaction with S₈in reﬂuxing toluene returned only starting

materials, and showed no new ³¹P NMR peaks, as ascertained

by measuring aliquots of the reaction mixture. Similarly, we

have not found any metals to which the phosphine coordi-

nates.

Whilst there is currently a strong interest in identifying

stable and persistent main group element free radicals,¹⁸there

is a paucity of phosphorus-containing examples.¹⁹Two-co-

ordinate R₂P^ꢂradicals are known, but the only stable deriva-

tive is the bis(trimethylsilyl)methyl derivative.²⁰A stable

diphosphanyl radical [Mes*(CH₃)P–PMes*]^ꢂhas also been

reported.²¹Stable phosphorus containing radicals that delo-

calize the unpaired electron over several nitrogen atoms are

also known; the 1,3-diphosphaallyl radical [ⁱPr₂NP(CNⁱPr₂)-

PNⁱPr₂]^ꢂ,²²the tetrakisamidophosphate radicals {Li₂[P(N^tBu)₃-

(NSiMe₃)]LiX ꢅ 3THF}^ꢂ(X = I, O^tBu),²³tris(dimethylhydra-

zino)diphosphine cation,²⁴several phosphaverdazyl radicals²⁵

and a vanadium stabilized NPN radical²⁶belong to this

category. In contrast, no stable radical cations of triarylphos-

phines have been reported before this work or that of Sasaki

Structure from X-ray crystallography

The molecular structure of Dipp₃P as determined by X-ray

crystallography is presented in Fig. 2, selected inter-atomic

distances and angles in Table 1, and full distances and angles

in Table S1, ESI.w All the crystals that we were able to prepare

through recrystallization from solvents ranging from alcohols

to aliphatic alkanes, as well as by vacuum sublimation, were

found to be twinned by merohedry.³⁴However, a twin law was

found which allowed for successful solution and reﬁnement of

the model (R₁= 2.94%).

ꢄc

This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientiﬁque 2008

New J. Chem., 2008, 32, 214–231 | 215

View Article Online

greater than those reported for Tripp₃P (P–C range from

1.839–1.851 A; +CPC range from 110.9(1)–111.9(1)1;¹²

P

{+CPC} = 334.411),¹⁴which crystallizes without crystal-

ꢀ

lographic site symmetry in P1. The slightly greater ﬂattening in

Dipp₃P vs. Tripp₃P is also substantiated by HF/6-31G(d)

P

calculations ( {+CPC} = 337.01 and 336.801, respec-

tively).¹⁴The P atom is found to be 0.43 A out of the least

squares plane deﬁned by C2, 3, 5, 6, which is more than in less

congested DippPR₂moieties, such as DippP(H)Si^tBuMe₂,

where the P is only 0.32 A out of the aromatic carbon atom

plane.¹¹

Fluorescence spectrum

Triphenylphosphine shows an unusually large ﬂuorescence

Stokes shift, which has been attributed to the occurrence of

a large geometrical change from pyramidal in the ground state

to near planar in the excited state.³⁵We therefore measured

the UV-Vis and ﬂuorescence spectra of Dipp₃P in anticipation

of observing a smaller Stokes shift because of its ﬂattened

ground state structure. The spectra (Table S2, ESI,w Fig. 3)

indeed demonstrate a much smaller Stokes shift in hexanes of

129 kJ mol^ꢀ1compared to 201 kJ mol^ꢀ1for Ph₃P.

NMR studies

Solution ³¹P NMR. The ³¹P chemical shift of Dipp₃P in

solution in CDCl₃is found at a high frequency for a tri-

arylphosphine, ꢀ49.7 ppm. For comparison, the shifts are ꢀ6

in Ph₃P, ꢀ35.3 in Xyl₃P, ꢀ37.2 in Mes₃P, ꢀ29.7 in Duryl₃P

and ꢀ53 ppm in the isosteric Tripp₃P.^12,36It has been shown

that +C–P–C angles are major contributors to the chemical

shifts of triorganophosphines, but larger angles are usually

associated with low-frequency chemical shifts.³⁷The compar-

ison with Tripp₃P indicates that the inductive electronic eﬀect

of the isopropyl substituents on the phosphorus shielding

tensor is not negligible.

Fig. 2 Thermal ellipsoid diagrams of the Dipp₃P molecule as found

in the crystal, (a) viewed down the crystallographic a axis and (b)

down the c axis. The P atom lies on a crystallographic threefold axis

and the symmetry-related Dipp rings are generated by the symmetry

operations (ꢀx + y, 1 ꢀ x, z) and (1 ꢀ y, 1 + x ꢀ y, z). H atoms are

omitted for clarity. Space-ﬁlling diagrams of these two views are

available in the preliminary communication.¹⁴

The converged model from the crystallographic study shows

the typical ‘‘propeller’’ conﬁguration of triaryl phosphines,

with the three aryl rings tilted to minimize interatomic repul-

sion. The molecular three-fold rotation axis aligns with the

crystallographic three-fold axis of the triclinic space group R3,

rendering the geometry of each of the Dipp groups identical.

The P–C bond length is 1.851(2) A, the C–P–C bond angle is

Solid state (SS ⁴) ³¹P NMR. The purpose in recording the

solid-state ³¹P NMR spectrum was to look for possible

magnetic inequivalence associated with twinning in the crys-

tals. The static solid-state ³¹P spectrum of Dipp₃P obtained

with direct polarization and proton decoupling has Dn_1/2

=

4.4 kHz, while the 15 kHz MAS spectrum under the same

conditions is a narrower singlet with Dn_1/2= 170 Hz and a

chemical shift of ꢀ49.5 ppm with respect to external 85%

111.88(5)1, and the sum of angles around phosphorus,

P

{+CPC}, is 335.641. Each of these parameters is slightly

Table 1 Selected interatomic distances (A) and angles (1) for Dipp₃P and Dipp₃P⁺from X-ray diﬀraction and computation^a

X-Ray diﬀr.

Dipp₃P

C₃in B3LYP/

6-31G(d)

D₃in B3LYP/

6-31G(d)

C₃in HF/

6-31G(d)

[Dipp₃P]⁺in UHF/

6-31G(d)

P–C1

C1–C6

C1–C2

P oop of C1 ꢃ 3^b

P oop C2–C3–C5–C6^c

1.8507(16)

1.418(2)

1.425(2)

0.539

1.8781

1.4236

1.4310

0.536

1.8142

1.4272

1.8777

1.4122

1.4221

0.531

1.8353

1.4164

1.4187

0.193

0

0.430

0.434

0.424

0.060

C1#1–P–C1

C6–C1–C2

C6–C1–P

C2–C1–P

a

111.88(5)

118.90(16)

127.17(13)

113.10(12)

112.2

118.8

126.8

113.4

120.0

121.4

112.3

118.5

127.2

113.4

118.9

121.3

122.5

116.2

119.3

b

Symmetry transformation used to generate equivalent atoms: #1 (ꢀx + y, 1ꢀx, z). Out of the plane of the three ipso C atoms of the aryl

c

rings. Out of the plane of the four ortho and meta C atoms of the aryl ring.

ꢄc

This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientiﬁque 2008

216 | New J. Chem., 2008, 32, 214–231

View Article Online

On cooling dilute solutions in either CDCl₃or (better)

CD₂Cl₂, the ¹H and ¹³C resonances go through smooth

coalescences and sharpen considerably by ꢀ80 1C, at which

point general line broadening limits further improvements in

the appearance of the spectra; fortunately at this temperature

the spectra are very clear. A low-temperature COSY spectrum

was measured to conﬁrm the assignment of the ¹H NMR

signals (ꢀ60 1C). The LT spectra of both nuclei deﬁne a static

structure in which the rotation of the isopropyl substituents is

frozen out and only methyl group rotation is fast on the NMR

timescale, and thus is consistent with molecular C₃symmetry

(Fig. S2, ESIw). There are in all four methyl and two methine

carbon environments. There are six diﬀerent ¹³C resonances in

the benzene ring. Most signiﬁcantly, the coupling to ³¹P

observed in the RT ¹³C spectra are found in each case to

become localized on nuclei belonging to one half of the Dipp

group at LT, consistent with conjugation to phosphorus in one

orientation, or of through-space interaction with the phos-

phorus lone pair. These couplings are typically about twice as

large at those observed in the averaged signals at RT.

Fig. 3 (a) Electronic absorption spectrum in hexanes (l_maxin nm;

log|e| in parentheses) and ﬂuorescence spectra of Dipp₃P in hexanes

(b) and ethanol (c). The ﬂuorescence intensity scale is arbitrary.

H₃PO₄. Cross-polarization (CP ⁴) results in very similar

spectra (Fig. S1, ESIw). The static SS ³¹P NMR spectrum

under CP conditions of Tripp₃P has been reported without

details.^6a

Solution ¹H and ¹³C NMR. The NMR parameters (see

Table 2) of the compound substantiate the putative molecular

structure, but are indicative of dynamic eﬀects because the RT

¹H and ¹³C NMR spectra (in both CDCl₃and CD₂Cl₂

solutions) reﬂect unexpectedly high molecular symmetry (i.e.

D₃, Fig. S2, ESIw). The para and meta aromatic proton signals

are distinctly resolved (as commonly seen in Dipp chemistry),

forming a pseudotriplet and pseudodoublet, but the latter is

further split by coupling to ³¹P. The isopropyl CH resonance is

found at 3.497 ppm and is a doublet of septets from coupling

to phosphorus and to the methyl protons. There are two

distinct isopropyl methyl resonances, one at 1.17 ppm, the

other at 0.71 ppm, and this latter is appreciably line broadened

at RT in both the 250 and 500 MHz spectra. This last

resonance is somewhat deshielded from the natural resonance

position of such methyl groups, which, based on the X-ray and

ab initio structure determinations and much previous experi-

ence, we attribute to the ‘‘paraphane’’ eﬀect, i.e. anisotropic

shielding induced when the protons on a methyl group are

swept across the surface of a neighbouring aromatic ring.¹⁰In

the ¹³C spectrum, two methyl carbon resonances (with the

lower ﬁeld signal dynamically broadened), one methine and

four from the benzene ring are observed at RT.

Dynamic NMR measurements. A thorough dynamic NMR

investigation of Dipp₃P using both ¹H and ¹³C spectra was

undertaken. The barriers for the DNMR processes were

determined using two methods, line shape analysis (LSA)

and the full width at half height method (Dn_1/2).³⁸The analyses

were undertaken by complete line ﬁtting for the methyl regions

1

in the H (Fig. 4) and ¹³C (Fig. 5) spectra where the rate and

Dn_1/2were determined as a function of temperature. The rates

were related to temperature using the Eyring relationship,

where ln(k/T) was plotted as a function of 1000/T (Fig.

S3–5, ESIw) to determine the activation enthalpy, DH^z, from

the slope and the activation entropy, DS^z, from the intercept.

(Merely using the temperature dependence of Dn_1/2, only an

upper bound to DH^zcan be obtained from the slope of an

‘‘Eyring like’’ plot of 1/Dn_1/2T vs. 1/T; however all information

on DS^zis lost.) This analysis is speciﬁc to uncoupled equally

populated AB systems, where the temperature dependence of

the frequency diﬀerence between A and B is expected to be

weak, as is the case for carbon and less so for proton. Table 3

lists the energetic results of the analysis, and the original plots

are provided in the ESI.w

Table 2 ¹H and ¹³C NMR data for Dipp₃P at ambient and low temperature^a

C₁

C₂

C₆

C₄

C₃

C₅

C₁₀

C₇

C₉

C₁₁

C₈

C₁₂

¹H: +30 1C, CD₂Cl₂

d

J

—

7.288

7.68

0

7.105

7.68

3.28

3.497

6.72

5.20

1.173

6.72

0

0.709

6.72

0

_HH/Hz

_HP/Hz

¹H: ꢀ80 1C, CD₂Cl₂

d

J

—

7.27

7.76

0

7.07

7.58

3.3

7.07

7.58

3.3

3.30

B6

0

3.14

B6

B6.4

1.13

6.1

0

0.99

6.3

0

0.90

6.3

0

0.36

6.2

0

_HH/Hz

_HP/Hz

¹³C: +20 1C, CDCl₃

d

J

135.37

25.9

153.64

18.6

129.38

0

124.37

3.9

32.37

17.6

32.37

17.6

23.30

0

24.82

0

_CP/Hz

¹³C: ꢀ80 1C, CD₂Cl₂

d

J

134.26

25

153.00

40

151.94

0

128.64

0

125

6

122.47

0

32.19

0

31.40

32

21.77

0

22.65

0

23.59

0

24.41

0

_CP/Hz

a

The atom numbering scheme is that used in Fig. 2; H atoms are those attached to the labeled C.

ꢄc

This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientiﬁque 2008

New J. Chem., 2008, 32, 214–231 | 217

View Article Online

Fig. 5 Line-shape ﬁtting for the isopropyl methyl group signals of the

¹³C NMR spectra of Dipp₃P as employed in (a) Line shape analysis

(LSA) and (b) full-width at half-height (Dn_1/2) analysis. The gray lines

represent the ﬁts to the experimental data.

Fig. 4 Line-shape ﬁtting for the isopropyl methyl group signals of the

¹H NMR spectra of Dipp₃P as employed in line shape analysis. The

gray lines represent the ﬁts to the experimental data.

theory only. Our attempts to optimize the structure of the

latter by DFT were not successful, but UB3LYP/6-31G(d)//

UHF/6-31G(d) calculations were performed at the HF opti-

mized geometry to determine the hyper-ﬁne coupling con-

stants (Table S3, Fig. S7, ESIw). The geometrical results

(Fig. 6) are presented in Tables 1 and S1, ESI,w and show

the very close agreement between the X-ray crystal and

ground-state calculated structures determined for Dipp₃P at

the two levels of theory. The only signiﬁcant deviation is in the

C–P bond lengths, which are longer in the calculated com-

pared to the experimental structure of the neutral molecule.

Note that the experimental C–P bond distance at 1.851(2) is

itself signiﬁcantly longer than the average value of 1.836(10) A

determined for 102 triaryl phosphines reported in the CCDC

database.⁴⁰The calculations also accurately reproduce the

distortion of the P atom out of the plane of four carbon atoms

of the aromatic rings (the two ortho and two meta atoms are

used to deﬁne these planes, see above). Another measure of a

pyramidal geometry is the deviation of the P atom from the

plane deﬁned by the three ipso carbon atoms of the Dipp rings.

Here too there is excellent agreement between the diﬀraction

results and the two computational methods.

The values obtained for the enthalpy of activation of the

dynamic process measured from the diﬀerent signals are

reasonably coherent. The results from the proton LSA indicate

that the enthalpy lies near 49 ꢁ 1 kJ mol^ꢀ1, but this value is

greatly underestimated by the Dn_1/2method, especially when

using ¹H NMR data. The lineshape analysis method was

augmented with data from selective inversion experiments at

the lowest four temperatures, which serves to correct for

systematic errors commonly encountered when relying on

LSA alone. This results in more accurate results but often

with higher statistical uncertainty.³⁹The carbon analyses agree

well within experimental error with the LSA proton results. A

free energy of activation for Tripp₃P has been reported as

46.0 kJ mol^ꢀ1at 243 K (in excellent agreement with our

results), which was also attributed to an inversion process,

1

but no details of the analysis are provided.^6aThe LSA on H

spectra indicates that the entropy of activation is positive with

a value of 24–27(ꢁ5) J mol^ꢀ1

K

^ꢀ1. The large positive activa-

tion entropy indicates an increase in the molar volume when

going to the transition state. It also indicates that the solvent

rearranges to a less ordered state when the molecule is

promoted from the ground to the transition state.

An estimate of the energy of a possible planar transition

state structure of Dipp₃P has been calculated by constraining

the molecule to D₃symmetry (Fig. 6(b)). The energy diﬀerence

between the C₃ground state and this planar geometry is 37.5

kJ mol^ꢀ1, implying a very low gas-phase barrier to pyramidal

inversion at the B3LYP/6-31G(d) level of theory. This planar

structure has P–C bonds that are 0.064 A shorter than in the

Quantum chemical calculations

We have obtained the structure of Dipp₃P in the C₃ground

and D₃transition states by both HF/6-31G(d) and B3LYP/6-

31G(d) methods and [Dipp₃P]⁺^ꢂat the UHF/6-31G(d) level of

ꢄc

This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientiﬁque 2008

218 | New J. Chem., 2008, 32, 214–231

View Article Online

Table 3 Energetic data from DNMR measurements for Dipp₃P

Proton

Carbon

Method

Methyl downﬁeld

Methyl upﬁeld

Methyl downﬁeld

1

2

Slope (from Dn_1/2plots)

ꢀ4.92 ꢁ 0.18

40.9 ꢁ 1.5

ꢀ5.01 ꢁ 0.19

41.7 ꢁ 1.6

ꢀ5.41 ꢁ 0.08

45.0 ꢁ 0.7

ꢀ5.58 ꢁ 0.18

46.4 ꢁ 1.5

Dn

Enthalpy/kJ mol^ꢀ1

Slope (from Eyring plots)

Enthalpy/kJ mol^ꢀ1

ꢀ5.89 ꢁ 0.09

48.98 ꢁ 0.69

24.8 ꢁ 4.3

ꢀ5.94 ꢁ 0.09

49.39 ꢁ 0.71

27.0 ꢁ 4.5

ꢀ6.35 ꢁ 0.36

52.80 ꢁ 3.0

44.0 ꢁ 19.7

ꢀ6.13 ꢁ 0.47

50.99 ꢁ 3.9

34.7 ꢁ 25.1

LSA

Entropy/J mol^ꢀ1K^ꢀ1

ground state, suggesting that the steric congestion of the bulky

Dipp groups is relaxed signiﬁcantly when the phosphorus

lone-pair becomes stereochemically inactive.

than for oxidation of Fc) on Au, Pt and GC electrodes in both

CH₃CN and CH₂Cl₂solvents (see Table 4 and S4, ESIw).

Comparison of characteristics such as larger peak-to-peak

separation, relative to that of known reversible one-electron

Cc^+/0and Fc^+/0processes measured under the same condi-

tions, implies that a quasi-reversible one-electron process:

The structure of Dipp₃P⁺at the UHF/6-31G(d) level of

ꢂ

theory is almost, but not entirely, planar at phosphorus, with

+C–P–C = 118.91 and a deviation of the P atom out of the

plane of the ipso carbon atoms of only 0.193 A (Fig. 6(d)). The

distortion of the P atom out of the plane of four carbon atoms

of the aryl ring is also greatly relieved at only 0.060 A. Thus

the single remaining non-bonded electron at phosphorus has

lost almost all of its stereochemical activity; this will have

important consequences for the EPR spectrum of the radical

cation (see below). In this species, the P–C bonds are also

considerably shorter than those in the neutral analog—

whether by comparison to the X-ray structure or to the HF/

6-31G(d) calculations—which is also indicative of the relaxa-

tion of steric pressure in the ion.

Dipp₃P " [Dipp₃P]⁺+ e^ꢀ

(1)

ꢂ

is detected under conditions of cyclic voltammetry. Addition-

ally, the presence of a small level of uncompensated resistance

in the form of Ohmic iR_uis indicated from the concentration

dependence in CH₂Cl₂. A second, irreversible, oxidation pro-

cess labeled as B in Fig. 7 is observed at a considerably more

positive potential just prior to the solvent limit in CH₂Cl₂.

Whether this is derived from further oxidation to P(^V), or an

oxidation at the electron-rich aryl ring, has not yet been

elucidated and this process is not discussed further. The almost

linear plots of the I_pvs. n^1/2obtained for the ﬁrst Dipp₃P

oxidation process in CH₃CN and CH₂Cl₂at GCE, Pt and Au

working electrodes imply that the mass transport process at

the peak potential is controlled by diﬀusion. From the

Randles–Sevcik relationship:⁴¹

Electrochemistry

Cyclic voltammetry

Dipp₃P undergoes an initial chemically reversible oxidation

process (Fig. 7) at modest potentials (slightly more positive

I_p= kn^2/3AD^1/2Cn^1/2

(2)

an estimation of the diﬀusion coeﬃcient was obtained from a

straight line ﬁt to the I_pvs. n^1/2plots. Results are summarized

Fig. 7 Cyclic voltammetry of Dipp₃P (0.20 mM) in CH₂Cl₂solution

containing 0.4 M ⁿBu₄NPF₆under an atmosphere of dry N₂recorded

at scan rate of 200 mV s^ꢀ1using an Au macrodisk electrode at 0 1C.

Fig. 6 B3LYP/6-31G(d) computed structures of (a) Dipp₃P in the C₃

ground state and (b) D₃planar transition state. (c) HF/6-31G(d)

computed structure of Dipp₃P and (d) UHF/6-31G(d) computed

structure of Dipp₃P⁺^ꢂ. The ‘‘tube’’ representations show only the P

(black) and C (gray) atoms.

Potentials vs. Fc^+/0. Note the wide window within which [Dipp₃P]⁺

ꢂ

is stable; the secondary process labeled ‘‘B’’ has not been investigated

in detail.

ꢄc

This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientiﬁque 2008

New J. Chem., 2008, 32, 214–231 | 219

View Article Online

Table 4 Reversible potentials (V)^aobtained for the [Dipp₃P]^+/0

process

CH₃CN^b

b

Solvent

CH₂Cl₂

E_f1 vs. Fc^+/0

0.09^c

1.44

0.66

0.79

1.25^d

0.18^c

1.53

0.70

0.84

E_f1 vs. Cc^+/0

E_f1 vs. Fc*^+/0

DE_f1 (Cc^+/0ꢀ Fc*^+/0

)

E_a(2nd process) vs. Fc^+/0

a

Calculated by cyclic voltammetry from the average of oxidation and

M

ⁿBu₄NPF₆at

b

reduction peak potentials. Containing 0.1

295 K. Cc^+/0data converted to the Fc^+/0scale by subtraction of

c

1.35 V. Scan rate = 0.2 V s^ꢀ1

.

d

in Table S5, ESI.w Excellent agreement was obtained between

digital simulation and experimental results for a quasi-rever-

sible one-electron oxidation using the parameters given in

Table 5 (Fig. 8). The estimate of the diﬀusion coeﬃcient from

simulation was in agreement with the values obtained inde-

pendently from steady state methods on micro-Pt and -carbon

electrodes and hydrodynamic measurements at a rotating disc

electrode (Table S5, ESIw).

Since the structure of Dipp₃P is close to spherical, and

presents a sheath of hydrocarbon to the surrounding environ-

ment, speciﬁc solvation interactions are expected to be mini-

mal. The similar electrochemical behavior in CH₃CN and

CH₂Cl₂can be rationalized on this basis, despite the fact that

these solvents are at opposite ends of the dielectric constant

scale. This also implies that the Stokes–Einstein equation,

which relates the radii and diﬀusion coeﬃcients of molecules,

ought to be approximately valid. Measurements of the dis-

tances between H nuclei at opposite ends of the molecule, and

adding 1.20 A to each H for the van der Waals radius, gives a

diameter perpendicular to the three-fold axis of 12.6 A and

parallel to the axis of 10.7 A, thereby describing a slightly

oblate spheroid with a dimensional ratio of 0.85. On the

reasonable assumption that the molecule is approximately

spherical, a simple average radius can be taken as 5.8 A.

Hence the diﬀusion coeﬃcient can be calculated from the

Stokes–Einstein equation:⁴²

Fig. 8 Comparison of cyclic voltammograms obtained for oxidation

of Dipp₃P (0.56 mM) at a scan rate of 100 mV s^ꢀ1at a 1.6 mm Pt disk

electrode in CH₂Cl₂(0.05 M ⁿBu₄PF₆) (—) and by simulation using

parameters provided in Table 5 (J).

value at 295 K for D_i= 1.0 ꢃ 10^ꢀ5cm²s^ꢀ1. This value is in

good agreement with the values for D measured for dilute

solutions of Dipp₃P in CH₃CN by RDE and steady-state

voltammetry. D for Mes₃P^+/0has been estimated in CH₃CN

to be D = 1.6 ꢃ 10^ꢀ5cm²s^{ꢀ1 44}

.

It is diﬃcult to compare potentials obtained in diﬀerent

solvents. However, assuming that the reversible potential for

the Cc^+/0process is independent of solvent, then there is a

diﬀerence of about 90 mV in the E_f1 of the [Dipp₃P]^+/0process

with respect to the Cc^+/0redox couple in changing from

CH₂Cl₂to CH₃CN (Table 4). If Cp*₂Fe is used as an internal

standard and the value assumed to be solvent independent,

then a smaller shift of about 60 mV is observed.⁴⁵Bond et al.

have shown that the diﬀerence between E_f1 (Fc^+/0) and E_f1

(Cc^+/0) is solvent independent (DE_f1 = 1.35 ꢁ 0.01 V), but

this does not conﬁrm that each couple is independent of

solvent.⁴⁶The observed solvent dependence could be due to

the reference scale independent approximation or diﬀerences

in ion pairing between [Dipp₃P]⁺and electrolyte in the two

ꢂ

D_i= k_BT/6pZr

(3)

solvents or a small level of solvent interaction. A study where

the ionic strength was varied signiﬁcantly from 0.05 to 0.50 M

(Table S6, ESIw) as was the concentration of Dipp₃P (0.5 to

2.8 mM) induced a range of reversible potentials of only

20 mV. Thus, neither ion-pairing nor solvent eﬀects are

considered to be large contributors to the reversible

[Dipp₃P]^+/0potentials.

where k_Bis the Boltzman constant, T the absolute tempera-

ture, Z the (normal) viscosity of the solution and r the radius of

the dissolved molecule. On the reasonable assumption that the

n

viscosity of a CH₃CN (0.1 M Bu₄NPF₆) solution is close to

that of the pure solvent (0.369 mN s m^ꢀ2at 20 1C),⁴³we get a

Broadening of the voltammetric wave shape and peak

potential shifts with increasing analyte concentrations are

typically assigned to the inﬂuence of uncompensated resistance

Table 5 Parameters obtained from digital simulation of a cyclic

voltammogram of Dipp₃P^a

n

Solvent Electrode 10^ꢀ5D/cm²s^ꢀ1

a

k_s/cm²s^ꢀ1R_measured/O

in CH₂Cl₂containing 0.05 to 0.5 M Bu₄NPF₆. However, a

close to linear scaling of the Faradaic current is observed

(Table S7, ESIw) with concentration of analyte in CH₂Cl₂

(0.5–5.0 mM) which implies that the inﬂuence of uncompen-

sated resistance on the voltammograms is minimal under these

high supporting electrolyte conditions. Furthermore, the line-

ar dependence of I_pvs. concentration of Dipp₃P at constant

scan rate provides an estimate of the diﬀusion coeﬃcient

CH₃CN Pt

GC

1.25

1.00

0.5 0.04

0.5 0.08

330

180

CH₂Cl₂Pt^b

GC

1.10

1.20

0.5 0.01

3100

1400

a

Determined by comparison of experimental data and simulated

b

cyclic voltammograms. With 0.05 M ⁿBu₄NPF₆supporting electro-

n

lyte; all other data obtained with 0.1 M Bu₄NPF₆.

(1.04 ꢃ 10^ꢀ5cm²

s

^ꢀ1) from the slope and Randles-Sevcik

ꢄc

This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientiﬁque 2008

220 | New J. Chem., 2008, 32, 214–231

View Article Online

relationship which is in excellent agreement with the other

on a 1.01 ꢃ 10^ꢀ3M solution of Dipp₃P in CH₂Cl₂containing

n

0.1 M Bu₄NPF₆, which was protected from the atmosphere

determinations reported in Table S5, ESI.w

by a blanket of N₂gas. An initial recording of the electronic

spectrum was undertaken before electrolysis (Fig. 11). During

oxidative electrolysis, continuous repetitive spectral scans

from 800 to 250 nm were recorded until such time that the

current was 1% of the initial value. Fig. 11 also presents key

spectra selected from the 38 recorded during forward electro-

lysis. The many isosbestic points are consistent with the direct

conversion of species Dipp₃P to [Dipp₃P]⁺^ꢂ. Furthermore, the

resulting solution of oxidized material was subsequently re-

duced at an applied potential held 0.4 V more negative than

the oxidation, resulting in almost complete conversion of the

spectroscopic signal back to that of Dipp₃P.

Fourier transform voltammetry

The signiﬁcant advantages of employing large amplitude ac

voltammetry in conjunction with Fourier reconstructions for

the analysis of electrode processes have been recently demon-

strated.¹⁷In a typical dc voltammetric experiment, it is not

necessarily obvious which terms (e.g. uncompensated resis-

tance, capacitance, k_S, E_f1, a) determine the shape and other

characteristics of the observed voltammograms. With the large

amplitude ac method and access to data over a wide frequency

range in a single experiment, the dc term is retained, but

numerous additional easily identiﬁable patterns of behavior

are now available from the power spectrum and the funda-

mental, second, third and higher harmonics. Each of these

terms exhibit characteristic dependencies on the above para-

meters that depend on concentration, frequency and scan rate,

but with diﬀerent levels of sensitivity, so that the prospect of

obtaining a unique solution to the electrode mechanism is

greatly enhanced.

The ﬁnal scan in Fig. 11 represents the best-characterized

UV-visible spectrum o_ꢂf a triarylphosphine radical cation. The

spectrum of Dipp₃P⁺is extremely rich (data in Table S2),

consisting of (at least) two broad bands in the visible region

(l_max: 498 and 456 nm), and several sharp bands in the UV

region (l_max: 373, 357, 341(sh), 296(sh), 285, 281(sh), 251 nm).

In this spectral region in CH₂Cl₂, Dipp₃P only has two bands

(l_max: 327 and 255 nm). Previously, only poorly resolved

Fig. 9 illustrates the data collected in a large amplitude ac

voltammetric experiment, displayed as a function of time.

Deconvolution into the dc component and ac harmonics

(Fig. 10) is achieved by a Fast Fourier Transform (FFT)–

Inverse Fast Fourier Transform (IFFT) sequence.^17fThe

patterns displayed in the harmonics presented in Fig. 10 are

fully consistent with both a quasi-reversible electron-transfer

process and uncompensated resistance. Importantly, increas-

ing the concentration of Dipp₃P from 0.50 mM up to 9.23 mM

leads to very characteristic changes in the third and higher

harmonics that can be well modeled by incorporating both

slow electron kinetics and uncompensated resistance. These

patterns cannot be simulated by considering uncompensated

resistance eﬀects or slow electron transfer alone.

spectra of Ar₃P⁺were known, produced in various matrices

ꢂ

by pulse radiolysis. These spectra show broad, ill-deﬁned

features with lowest energy maxima of ca. B500–530 nm.⁴⁸

The qualitative appearance of the spectra reported for Ph₃P⁺

ꢂ

and (p-Tol)₃P⁺^ꢂat 77 K in n-butyl chloride matrices, however,

are in agreement with our results.⁴⁹It seems likely that the rich

spectral features of the radical cation reﬂect transitions be-

tween the remaining phosphorus non-bonded electron and the

many p* levels of the electron rich Dipp aryl rings. The visible

bands account for the distinct orange color of dilute solutions

of Dipp₃P⁺salts. Ph₃N⁺has an intense absorption at

640 nm attributed to delocalization of the unpaired electron

over the aryl rings.⁵⁰The absence of such a very-low energy

band, as well as evidence from EPR spectra (see below), is

against the hypothesis of extensive delocalization of this kind

ꢂ

Spectroelectrochemistry

in Dipp₃P⁺

.

ꢂ

Optically transparent thin-layer electrolysis (OTTLE) spectro-

scopic⁴⁷monitoring of the oxidation of Dipp₃P was conducted

Bulk electrolysis

The course of oxidative bulk electrolysis of Dipp₃P in CH₂Cl₂

(0.1 M ⁿBu₄NPF₆at 200 mV positive of E_f1) at a Pt gauze

basket electrode was monitored by obtaining rotating disk

electrode (RDE) voltammograms before and after exhaustive

oxidative electrolysis (Fig. 12 and S6, ESIw). The identical

magnitude, but opposite sign of the limiting currents conﬁrms

that Dipp₃P is quantitatively oxidized to [Dipp₃P]⁺^ꢂ. Quanti-

tative conversion back to Dipp₃P was also achieved. Coulo-

metric measurements conﬁrm that the oxidation process and

reduction back to starting material both involve the transfer of

one electron (n = 0.88) per molecule. The radical cation

[Dipp₃P]⁺produced was shown to be stable for at least 2 h

ꢂ

when generated by bulk electrolysis.

Fig. 9 I–t data obtained from a large amplitude ac voltammogram of

Dipp₃P (0.5 mM) in CH₂Cl₂(0.51 M ⁿBu₄NPF₆) solution at a 0.94 mm

diameter Pt disk electrode; n = 47.68 mV s^ꢀ1, freq. = 21.46 Hz, amp

EPR spectra

Samples of [Dipp₃P]⁺were ﬁrst obtained by bulk electrolysis

in CH₂Cl₂(0.1 M ⁿBu₄NPF₆). In one case, the solution at the

ꢂ

= 80 mV. Data obtained over the potential range of 0–1 V vs. Ag^+/0

,

65 536 data points collected.

end of exhaustive electrolysis was transferred, with full

ꢄc

This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientiﬁque 2008

New J. Chem., 2008, 32, 214–231 | 221

View Article Online

Fig. 10 DC, fundamental and ﬁrst four harmonic components obtained from data presented in Fig. 9 (—) using a FFT-IFFT sequence and the

best-ﬁt simulated components (--). Parameters: D₀= D_R= 1.04 ꢃ 10^ꢀ5cm s^ꢀ1, a = 0.5, k = 0.01 cm s^ꢀ1, R_meas= 2100 O.

ꢂ

protection from the atmosphere, to a previously dried and N₂

ﬁlled (vacuum line) quartz EPR tube. In a second, a ten-fold

excess of Dipp₃P was added to the electrolyzed solution to

create an approximately 1 : 10 concentration ratio of

[Dipp₃P]⁺and Dipp₃P. The spectra were recorded as solu-

tions at 22 1C and in a frozen glass at ꢀ143 1C (Fig. 13). The

isotropic liquid spectra show the presence of the expected

doublet with a hyperﬁne coupling to ³¹P of 23.9 mT (Table 6).

The peak-to-peak linewidths of the spectral lines was found

1

between 0.5 and 0.6 mT. Since H is the only other magneti-

cally active nucleus of signiﬁcant abundance in the structure, it

is likely that the wide lines are at least in part due to

Fig. 11 Oxidation of 1.01 mM Dipp₃P in CH₂Cl₂(0.1 M ⁿBu₄NPF₆)

under OTTLE conditions. The initial spectrum (*) with a maximum at

375 nm is from pure Dipp₃P (see Fig. 3 and Table S2, ESIw). Decrease

(k) of this band and growth (m) of many new bands are indicated by

arrows. The ﬁnal spectrum with maxima at 251, 285, 357, 373, 456 and

498 nm (among others) i_ꢂs the spectrum of essentially pure [Dipp₃P]⁺

Fig. 12 RDE curves at a GC electrode before (upper) and after

(lower) bulk electrolysis when Dipp₃P⁺is generated from 8.6 mM

Dipp₃P in CH₂Cl₂(0.5 M ⁿBu₄NPF₆). For simulations of RDE

voltammograms see Fig. S6, ESI.w

ꢂ

.

Reduction of [Dipp₃P]⁺back to [Dipp₃P] in the OTTLE cell reverses

the sequence of spectra while retaining the isosbestic points.

ꢄc

This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientiﬁque 2008

222 | New J. Chem., 2008, 32, 214–231

View Article Online

the time scale of bulk electrolysis (i.e. hours in CH₂Cl₂

solution at room temperature) suggested that chemical oxida-

tion of Dipp₃P should be possible. The reversible potential of

the [Dipp₃P]^+/0couple is at +0.09 V vs. Fc^+/0, and no further

oxidation occurs until +1.25 V vs. Fc^+/0. The list of potential

chemical oxidizing agents provided by Connelly and Geiger⁵⁵

advises that Ag(I) ion in CH₂Cl₂solution should be a suitable

oxidant at the Ag^+/0redox couple of +0.65 V vs. Fc^{+/0 56}

.

This is an attractive reagent because the by-product is ex-

pected to precipitate as solid silver. We have therefore used

AgClO₄, AgPF₆and AgSbF₆as chemical oxidants to generate

solutions of [Dipp₃P⁺^ꢂ]X^ꢀ(X^ꢀ= ClO₄^ꢀ, PF₆^ꢀ, SbF₆^ꢀ) salts

in CH₂Cl₂. EPR spectra of dilute solutions of such chemically

generated radicals are identical to those prepared by electro-

chemical methods.

Discussion

Dipp₃P

Fig. 13 Comparison of simulated (J) and experimental (—) EPR

spectra obtained (a) in the glass state by freezing an electrochemically

generated 1 mM solution of Dipp₃P⁺in CH₂Cl₂(0.1 M ⁿBuN₄PF₆) at

130 K, and (b) at 295 K in CH₂Cl₂using a solution prepared by

AgClO₄oxidation of Dipp₃P. Simulations performed with WinEPR

software; for (a) 40% Lorentzian–60% Gaussian lineshape, LW> =

1.0, LWJ = 1.1 mT; for (b) with Lorentzian lines, LW = 0.12 mT and

Structure. Since the ﬁrst X-ray diﬀraction structure deter-

mination of Mes₃P by Mislow and co-workers,⁵⁷it has been

recognized that bulky substituents cause substantial distor-

tions in the structure of the putative pyramidal R₃P group.

Thus while PH₃(cone angle 871) has internal angles of 93.61,⁵⁸

Mes₃P has greatly enlarged internal angles (average

109.71).^36b,57The small bond angles of PH₃have long intri-

gued chemists and theoreticians. Molecular orbital theory has

been used to advance a convincing explanation in the form of

Walsh diagrams (Fig. S8, ESIw) that correlate the planar D_3h

geometry for EH₃molecules with the strongly pyramidalized

C_3vgeometry such as PH₃adopts.^51b,59Pyramidalization

occurs when the 2a₁orbital (labeled as in the lower-symmetry

point group; an empty p_zatomic orbital of E, for example, in

planar AlH₃, and the ‘‘lone pair’’ orbital of EH₃, for example

in PH₃) is occupied, because rehybridization of this orbital

allows for partial bonding to the s orbitals of the three

substituents at E. The stabilization of 2a₁often provides the

main energetic driving force favoring the pyramidal geometry,

while 1e is destabilized to a lesser extent. In this analysis, the

much greater pyramidalization of PH₃compared to NH₃can

be traced to the lower electronegativity of the third row

element, and hence reﬂects the fundamental identities of the

central elements N and P. When substituents of high steric

1

additional hfc to 6 H = 0.17 mT and 6 ¹H = 0.10 mT.

unresolved proton hyperﬁne coupling.^36a,51Using the program

WinSim,⁵²it was possible to get a superb ﬁt using purely

Lorentzian lines (as expected if all coupling is accounted for)

by including two sets of six equivalent ¹H hfc of 0.17 mT

(possibly isopropyl methine H) and 0.10 mT (possibly the meta

aromatic H) and a residual LW of only 0.12 mT.⁵³No

measurable diﬀerence in the line width was observed between

the two s_ꢂamples that contained on the one hand 100%

[Dipp₃P]⁺and the 1:10 mixture of [Dipp₃P]⁺^ꢂ/Dipp₃P,

indicating that line broadening through intermolecular elec-

tron transfer is not signiﬁcant for this system. In the frozen

glass, a characteristic axially-symmetric powder pattern was

obtained with parallel and perpendicular hfc tensor compo-

nents of 12.7 and 42.6 mT, and a small asymmetry in the g

tensor.^12,36,54

The successful_ꢂelectrochemical generation of solutions con-

taining Dipp₃P⁺and the demonstration of their stability on

Table 6 EPR parameters for Ar₃P⁺radicals

ꢂ

Compound

Conditions

g_iso

a_iso/mT

g_J

a_J/mT

g_>

a_>/mT

Dipp₃P⁺

CH₃CN/295 K (in situ)^a

CH₂Cl₂/295 K^b

CH₂Cl₂/77 K^c

2.008

2.007

2.008

2.007

2.0052

2.006

23.9

22.7

23.9

23.7

24.4

24.0

29.9

ꢂ

2.0045

2.002

42.6

2.0085

2.009

12.7

CH₂Cl₂/295 K^d

CH₂Cl₂/293 K^e

C₃H₇CN/77 K^f

CH₃CN/123 K^f

C₃H₇CN/77 K^g

CFCl₃/77 K^f

Tripp₃P⁺

41.7

42.0

41.1

40.2

45.2

13.0

15.0

17.0

17.1

22.3

ꢂ

Duryl₃P⁺

ꢂ

Xyl₃P⁺

ꢂ

Mes₃P⁺_ꢂ

Ph₃P⁺

a

ꢂ

Saturated solution (between 1 ꢃ 10^ꢀ4and 1 ꢃ 10^ꢀ3M) in CH₃CN containing 0.1 M ⁿBu₄NPF₆at 295 K. LW_Lo0.57; LW_Hi0.60 mT. 7.7 ꢃ 10^ꢀ4

b

M in CH₂Cl₂containing 0.1 M ⁿBu₄NPF₆at room temperature. LW_Lo0.54; LW_Hi0.59 mT. Sample as in b but cooled to 130 K; g = (g_J+ 2g_>)/

c

3; a = (a_J+ 2a_>)/3.^{54c d}Prepared from a dilute solution of Dipp₃P in CH₂Cl₂by chemical oxidation with solid AgClO₄. LW_Lo0.57; LW_Hi

e

0.60 mT. Chemical oxidation with AgClO₄.^{12 f}Prepared by electrolysis.^{54d g}Electrolysis in n-butyronitrile.^36a

ꢄc

This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientiﬁque 2008

New J. Chem., 2008, 32, 214–231 | 223

View Article Online

pressure are placed on R₃P, orbital hybridization will resist the

tendency towards ﬂattening. These opposing tendencies cause

inter alia the strong displacement of the P atom out of the

plane of the aryl rings, as well as a similar distortion of the ipso

C. Thus there seems to be a limit to the amount of planariza-

tion that can be imposed at P by bulky (carbon-based)

substituents.⁶⁰

the gas phase. For more detailed insight into this phenomenon

in condensed phases, we turn now to the solution dynamics as

measured by NMR.

Solution dynamics. Slow dynamics processes in crowded

triarylphosphines have been investigated previously by NMR

methods.^5b,63–65Rieker and Kessler,⁶³supported later by

Mislow and co-workers,⁵⁷argued that the correct mechanism

for exchange in Mes₃P is by a ‘‘two-ring ﬂip’’ mechanism, a

kind of turnstile-motion in which the rotation of one of the

three aryl rings induces a ﬂip of the remaining two to achieve

the opposite propeller rotation. Stepanov et al. obtained a

lower activation energy in Duryl₃P, and used this as an

argument that the latter phosphine undergoes pyramidal in-

version instead of ring-ﬂipping.⁶⁴Bellamy et al. have con-

ﬁrmed the value of about 44.5–45.6 kJ mol^ꢀ1for the propeller

reversal in Mes₃P.⁶⁵Based on Kessler’s concept of measuring

the steric interactions between the ortho groups in congested

propeller molecules of this type, it would be expected that the

Solid-state NMR. When Mes₃P is crystallized from ethanol/

toluene, a polymorph in the space group Cc is obtained in

which the phosphorus atom positions are disordered by an

undetermined mechanism.^36bRecrystallization from CHCl₃

ꢀ

produces a diﬀerent polymorph (space group P1) which has

two independent molecules in the unit cell. Penner and Wasy-

lishen detected the polymorphism of Mes₃P through the

discovery of three ³¹P resonances in the CP-MAS spectra of

commercial Mes₃P.⁶¹Recrystallization from chloroform re-

moved the broadest (Dn_1/2= 49 Hz) of these lines and left

behind two very sharp (Dn_1/2= 16 Hz) resonances in a 1 : 1

ratio which are ascribed to the two independent molecules in

i

considerably larger Pr groups in Dipp₃P and Tripp₃P should

ꢀ

the asymmetric unit of the P1 polymorph, in which disorder

lead to a signiﬁcantly higher barrier for P–C_ipsobond rotation,

was not encountered. These workers showed that asymmetry

in the shielding tensors of triaryl phosphines is caused pri-

marily by the conformations of the aryl rings in the solid state.

Where these are twisted in diﬀerent orientations, or where they

have substituents in exo as well as endo orientations, the ³¹P

shielding tensors are signiﬁcantly asymmetric. Those phos-

phines that crystallize with the rings in similar (or crystal-

lographically identical) conformations and those that have all

exo orientations of ortho alkyl groups tend to have the most

symmetric shielding tensors. The conformation of the three

Dipp rings is identical in Dipp₃P by crystallographic symmetry

in space group R3, so that the shielding tensor in absence of

other phenomena ought accordingly to be cylindrically sym-

whereas we have measured barriers by DNMR of only

49 ꢁ 1 kJ mol^ꢀ1

.

In an elegant study on the quaternary cation [Mes₃PCH₃]⁺

in presence of a chiral resolving agent, Laleu et al. have

recently provided direct experimental evidence for the two

ring-ﬂip mechanism.⁹However, this approach is not possible

for neutral Dipp₃P, as such chiral recognition depends on

ionic charges. We have instead undertaken a detailed ab initio

computational study in which one of the Dipp groups was

forced to rotate about the P–C1 bond with respect to dihedral

angle y (C1⁰–P–C1–C2) as deﬁned in Scheme 1, while one

other dihedral angle f (C1–P–C1⁰⁰–C2⁰⁰) was held constant,

and the remainder of the molecule was allowed to relax. From

a full investigation of the energy surface resulting from varying

y in steps for diﬀerent f, the only low-energy path that

interconverts species appears to be pyramidal inversion

(Fig. 14).

metric. That the spectra obtained have lines wider (Dn_1/2

=

170 Hz) than those of the monoclinic (disordered), and much

wider than the triclinic phase of Mes₃P, is thus likely a

reﬂection of the twinning by merohedry (see Structure from

X-ray crystallography, above).

The dark blue zones in Fig. 14 correspond to two equal

energy C₃conformations (y = f = ꢀ831 and ꢀ191) which are

related to each other by pyramidal inversion without racemi-

zation (an equivalent energy surface exists for the enantio-

meric structure which we have not calculated in detail.) The

red zones represent a very high energy wall surrounding the

whole enantiomeric surface. The process starts out as a turn-

stile mechanism, where the neighboring rings are made to

rotate in opposite directions, but at a particular point along

Photophysics. Triphenylphosphine is remarkable for the

unexpectedly large Stokes shift that is observed in its ﬂuores-

cence spectrum (Table S2, ESIw).³⁵Changenet et al. have

argued that this Stokes shift of B225 kJ mol^ꢀ1in ethanol

reﬂects its high barrier to pyramidal inversion because the

Ph₃P^zexcited state geometry is expected to be planar or close

to planar.^35aThus, in Ph₃PQO which cannot go planar, the

Stokes shift in ethanol is only 29 kJ mol^ꢀ1and on the

reasonable assumption that Ph₃P has a quite similar electronic

structure to that of its oxide, an upper limit for its pyramida-

lization energy was found to be o200 kJ mol^{ꢀ1 36b}

.

These

estimates can be compared to a barrier for pyramidal inversion

of 147 kJ mol^ꢀ1obtained from HF/6-31G(d)/6-31G ab initio

molecular orbital calculations.⁶²In the case of Dipp₃P, there is

no corresponding oxide available to index the Stokes shift as

done for Ph₃P; thus we can only conclude that the barrier to

pyramidal inversion is expected to be o100 kJ mol^ꢀ1in

hexane solution. Using DFT methods, we have calculated a

barrier to a planar ‘‘transition state’’ of only 37.5 kJ mol^ꢀ1in

Scheme 1

ꢄc

This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientiﬁque 2008

224 | New J. Chem., 2008, 32, 214–231

View Article Online

Tordo and co-workers have determined fast electron transfer

kinetics (k_S4 1 cm s^ꢀ1) for the Mes₃P^0/+couple.⁴⁴If this is

correct, re-organizational geometry changes upon oxidation

cannot be a cause of the slower electron transfer in Dipp₃P, as

even greater changes are predicted for Mes₃P⁺^ꢂ. A possible

origin for slower electron transfer in the Dipp₃P^+/0redox

couple is, we suggest, the encapsulating nature of the isopropyl

groups surrounding phosphorus, leaving a narrow channel

that aﬀords the only access to the central P atom. Moreover,

at RT in solution, the DNMR results just discussed suggest

rapid pyramidal inversion, which has the eﬀect of blocking

even further the access (or egress) of the electron from the P

atom. An idea of this can be obtained from a diagram taken of

the DFT computed planar inversion state, which is shown in

Fig. 16. In EPR studies of 10% Dipp₃P⁺in presence of

ꢂ

unoxidized Dipp₃P we found no evidence of line broadening

ascribed to intermolecular electron transfer. This lends addi-

tional support to the notion of rather slow electron transfer.

Fig. 14 Interpolated surface of the potential energy of Dipp₃P as a

function of the dihedral angles y and f (see text for deﬁnitions).

the way, a severe steric clash between the isopropyl groups of

neighboring Dipp rings prevents further rotation. This clash

apparently has a much larger energy demand than phosphor-

ous inversion, which subsequently follows instead of complet-

ing the full rotation. The trajectory followed by the molecule

along the lowest-energy pathway (Fig. 15) between the two

minima passes through a saddle-point corresponding to a

planar transition state (y = f = ꢀ561); hence, rotation

induces inversion⁶⁶in very bulky triaryl phosphines. The

calculated barriers to inversion of 54 kJ mol^ꢀ1at the HF/3-

21G and 37.5 kJ mol^ꢀ1at B3LYP/6-31G(d) levels of theory

are in reasonable agreement with the 51–53 kJ mol^ꢀ1obtained

from the DNMR experiments. Thus when steric pressure is

severe, the normal resistance to inversion at phosphorus can

apparently be overcome even with carbon-based substitu-

ents.⁶⁰

Comparison of phosphine redox potentials. There is a con-

siderable body of literature on the redox chemistry or

triarylphosphines.^{7h,12,49,54,67}Triphenylphosphine is irreversi-

bly oxidized at a potential of +1.04 V (anodic peak potential

vs. Fc^+/0) in CH₃CN (our data). A small selection of the many

previously reported potential values is presented in Table 7. It

becomes immediately obvious that serious problems asso-

ciated with use of diﬀerent reference potential scales aﬀect

the published data. We have therefore re-measured the com-

pounds listed in Table 7 in both CH₃CN and CH₂Cl₂solutions

using a common internal standard: the Cc^+/0couple;⁴⁶and

quote our data on the Fc scale as recommended by IUPAC.⁶⁸

In our hands, the Dipp₃P^+/0couple in CH₂Cl₂solution is

found to be 0.15 V less cathodic than the isosteric Tripp₃P^+/0

.

This diﬀerence is likely due to the inductive inﬂuence of the

i

three Pr groups in the para position in Tripp₃P that are

i

missing in Dipp₃P, or approximately 0.05 V per Pr group.

Electrochemistry. The exhaustive studies presented above

conﬁrm that Dipp₃P in both CH₃CN and CH₂Cl₂displays a

chemically reversible one-electron oxidation which is, how-

ever, marked by somewhat slow electron transfer kinetics.

Similarly the potentials of Xyl₃P^+/0and Mes₃P^+/0diﬀer by

0.15 V, thus B0.05 V per methyl group. Such ‘‘voltammetric

Fig. 16 View perpendicular to the three-fold rotation axis of the

B3LYP/6-31G(d) calculated structure of the ‘‘transition state’’ geo-

metry held to D₃point symmetry. The van der Waals radii scale is the

same as that used in the preliminary communication.¹⁴The phos-

phorus surface is rendered in black.

Fig. 15 A reaction coordinate (in arbitrary units) showing the lowest-

energy path connecting the degenerate ground state via the planar

transition state. The dihedral angles y corresponding to minima and

maximum are indicated.

ꢄc

This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientiﬁque 2008

New J. Chem., 2008, 32, 214–231 | 225

View Article Online

Table 7 Comparative redox potentials for triaryl phosphines^a

P

b

Compound

CH₂Cl₂

CH₃CN^c

Tordo^d

Sasaki^e

{+CPC}^f

E(HOMO)^f/eV

Ph₃P

(+0.88)

+0.30

+0.34

+0.19

+0.09

ꢀ0.06

(+1.04)

(+0.37)

+0.40

+0.26

+0.18

+0.04

(+1.400)

—

+0.885

+0.784

—

+1.03

—

+0.41

—

+0.16

310.00

334.10

330.51

330.33

337.01

336.80

ꢀ8.00

ꢀ6.67

ꢀ7.31

ꢀ7.20

ꢀ7.29

ꢀ7.13

Anthr₃P

Xyl₃P

Mes₃P

Dipp₃P

Tripp₃P

a

—

Calculated by cyclic voltammetry from the average of oxidation and reduction peak potentials, scan rate = 0.2 V s^ꢀ1; values in parenthesis are

E^oxpeak potentials. In CH₂Cl₂at Pt a electrode containing 0.4 M in ⁿBu₄NPF₆and 1 mM in analyte, vs. Fc^+/0, measured against Cc^+/0

,

b

expressed vs. Fc^+/0by subtracting 1.35 V.^{46 c}In CH₃CN at a Pt electrode containing 0.1 M in ⁿBu₄NPF₆and 1 mM in analyte, measured against

Cc^+/0, expressed vs. Fc^+/0by subtracting 1.35 V.^{46 d}In CH₃CH₂CH₂CN 0.1 M in ⁿBu₄NPF₆and 1 mM in analyte, vs. SCE.^{36a e}In CH₂Cl₂0.1 M

in Bu₄ClO₄measured against a pseudo-Ag^+/0reference electrode.^{12 f}From HF/6-31G(d) calculations.¹⁴

n

inductive inﬂuences’’ are well known in metallocene chemis-

try.⁶⁹Using a very extensive range of electrochemical mea-

surements, Tordo and co-workers demonstrated a distinct

inﬂuence on triarylphosphine oxidation potentials by the

degree of pyramidalization⁷⁰at phosphorus through a careful

consideration of isosteric species and by eliminating the

inductive eﬀects of substituents.^36a,51bAt the same time, these

workers have pointed out the signiﬁcance of substituent

inductive eﬀects and showed that good correlations with

Hammett parameters could be obtained when the phosphines

with, and those without, para substituents are treated

separately.^51bSince the inductive eﬀects on redox potentials

of ⁱPr and Me groups in this system are the same, the

signiﬁcant Mes/Tripp (ꢀ0.25 V) and Xyl/Dipp (ꢀ0.25 V)

shifts do seem to be mainly due to the structural changes

induced by the ortho ⁱPr groups, but whether this is truly from

planarization or from activation by distortion is diﬃcult to

decide (see under Structure of Dipp₃P above).⁷¹

Indeed, Anthr₃PQO is a known compound,⁷⁴whereas neither

Dipp₃P nor Tripp₃P react with oxygen in the form of O₂or H₂O₂.

Spectroscopy. The EPR parameters for Dipp₃P⁺(Table 6)

ꢂ

may be compared to those of previously prepared stable and

unstable R₃P⁺^ꢂ. All the parameters are similar to, but not

identical with, those of the isosteric Tripp₃P⁺^ꢂ, suggesting that

inductive eﬀects from the alkyl groups aﬀect the electronic

properties of the radicals as well.¹⁴Our value for, in particular,

the a_>component of the hfc tensor is the smallest of the series

presented in Table 6, and may be compared to the values

recorded for Ph₃P⁺^ꢂ, Mes₃P⁺^ꢂ, Xyl₃P⁺^ꢂ, Duryl₃P⁺and

ꢂ

67

Tripp₃P⁺

.

Tordo has argued that this parameter indicates

ꢂ

the degree of planarization of the phosphorus pyramid in the

radical cation.^67d,eOur gas-phase calculation on Dipp₃P⁺

ꢂ

gives a +(CPC) value of 118.91 (Table 1), which corresponds

to Tordo’s ‘‘a’’ = 6.01.⁷⁰Thus the evidence from EPR

spectroscopy and calculations is that excessively bulky sub-

stituents such as Dipp and Tripp cause signiﬁcant ﬂattening in

Ar₃P⁺

.

ꢂ

[Dipp₃P⁺^ꢂ] radical cation

Stability. Extensive studies have shown that the end product

of electrochemical oxidation of Ph₃P is always Ph₃PQO.⁷²

More important to this study is the fact that a detailed

investigation has shown that a similar process operates for

Mes₃P, where in wet or oxygenated solvents, Mes₃PQO is

formed, likely along with Mes₃PH⁺.^44,73By contrast, solu-

tions of Dipp₃P prepared in CH₃CN without special drying

procedures and without purging by N₂or Ar, display essen-

tially identical voltammetric behavior in the 0/+1 process on

the timescale of CV or RDE to that observed for carefully

dried and deoxygenated solutions, although preparations of

Conclusions

Triarylphosphines with two ortho ⁱPr substituents per aryl ring

are currently the most sterically congested phosphines known.

They have unusual properties which include high thermal

stability and volatility, stereochemical non-rigidity apparently

due to the rare phenomenon of pyramidal inversion at phos-

phorus, and facile electrochemical oxidation. Th_ꢂeir radical

cations are extremely stable—samples of Dipp₃P⁺prepared

in sealed vessels under rigorously dry conditions retain their

EPR intensity for months—which is attributable primarily to

the encapsulating eﬀect of the bulky ortho isopropyl groups.

Work in our laboratories is ongoing to delineate the reactivity

of Dipp₃P and to isolate stable salts of its radical cation. Thus

far, all salts of Dipp₃P⁺that we have been able to prepare

appear to be amorphous to X-rays.

Dipp₃P⁺exposed to the atmosphere do degrade on the

ꢂ

timescale of hours to as yet unknown products. We attribute

this enhanced stability of the radical cation to the superior

steric shielding aﬀorded by the large enveloping exo ⁱPr

groups. Support for this assertion is provided by the electro-

chemically and chemically irreversible oxidation of Anthr₃P

(Table 7), at considerably higher potentials, despite the ap-

Experimental

proximately equal steric pressure of the Anthr groups com-

P

General methods

pared to Dipp (consideration of

+(CPC) from both

crystallography and computation).¹⁴The Anthr substituent

does not provide adequate steric shielding for the phosphorus

lone pair, and its electrochemical behavior is much closer to

that of Ph₃P than any of Xyl₃P, Mes₃P, Dipp₃P or Tripp₃P.

DippBr was prepared as previously reported.⁷⁵PCl₃, anhy-

drous CuCl, Mg turnings and silver salts (Aldrich) were used

as received. THF was distilled from sodium benzophenone

immediately before use. n-Heptane and xylenes were distilled

ꢄc

This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientiﬁque 2008

226 | New J. Chem., 2008, 32, 214–231

View Article Online

from LiAlH₄and Na, respectively, under N₂. Hexanes, chloro-

form and all deuterated solvents were used as received. All

reactions were performed in an atmosphere of dry N₂in oven-

dried glassware (44 h at 4160 1C) unless otherwise noted.

were included at geometrically idealized positions and were

not reﬁned. The ﬁnal cycles of full-matrix least-squares reﬁne-

ment using SHELXTL⁷⁹converged. The weighting scheme

was based on counting statistics and the ﬁnal diﬀerence maps

are essentially featureless.

Dipp₃P

CCDC reference number 660828.

For crystallographic data in CIF or other electronic format

see DOI: 10.1039/b709602j

Synthesis. To a mixture of magnesium turnings (0.56 g, 22.5

mmol), I₂(50 mg, 0.2 mmol) and tetrahydrofuran (50 mL) was

added DippBr (5.41 g, 22.5 mmol), and the mixture was stirred

over night at room temperature. The resultant solution of

DippMgBr⁷⁵was cooled to ꢀ70 1C and CuCl (98%, 2.27 g,

22.5 mmol) was added under N₂. The mixture was stirred for

about 1 h at ꢀ70 1C, then warmed to RT and stirred over-

night, evaporated, re-dissolved in 50 mL freshly distilled

n-heptane, the suspension cooled to ꢀ70 1C and PCl₃(98%,

1.05 g, 0.67 mL, 7.64 mmol) added. The mixture was stirred

for about 45 min at ꢀ70 1C, warmed to RT and heated to

reﬂux for 40 h then cooled to RT. Hexanes (30 mL) was added

to the mixture and the resultant suspension was ﬁltered

through Celite with suction. Evaporation of solvent left a

sticky solid, which could be recrystallized from hot hexanes–

chloroform to aﬀord crystalline Dipp₃P (2.33 g, 4.52 mmol,

B60% yield).⁷⁶Calc. for C₃₆H₅₁P: C, 84.00; H, 9.99%.

Found: C, 84.26; H, 9.71%. Mass spectrometry (EI):

514.3744 (C₃₆H₅₁P⁺, ꢀ1.1 ppm, 100%); 471.31757

(C₃₃H₄₄P⁺, +1.0 ppm, 9%). mp: sublimes o360 1C/1 bar.

NMR data: ¹H, ¹³C see Table 2; ³¹P d ꢀ49.7 w.r.t. ext. H₃PO₄.

Luminescence spectroscopy. The spectrum of Dipp₃P was

obtained in ﬂuorescence mode using 5.8 ꢃ 10^ꢀ4M solutions in

cyclohexane and saturated solutions in 95% EtOH. A Perkin

Elmer LS50B luminescence spectrometer was employed for the

measurements, which are displayed on an arbitrary intensity

scale in Fig. 3 (numerical data in Table S2, ESIw).

NMR spectroscopy. NMR spectra were obtained on Bruker/

Tecmag AC250 (¹H, ¹³C and ³¹P) and Varian Innova 500

spectrometers (VT ¹H and ¹³C). ¹H and ¹³C spectra are

referenced to (residual) solvent peaks, while ³¹P NMR spectra

were referenced to external 85% H₃PO₄and collected either in

locked mode for the puriﬁed ﬁnal product or in unlocked

mode when aliquots of the reactions were investigated to

monitor the completeness of the reactions.

For the dynamic NMR studies, solutions of Dipp₃P were

prepared in CD₂Cl₂at 10% w/v which were transferred to a 5

mm thin walled NMR tube. The low temperatures were

attained using liquid N₂as the coolant, and the output

temperatures for the thermocouple were calibrated against

the chemical shift diﬀerence of methanol over the appropriate

temperature range.⁸⁰Both ¹H and {¹H}–¹³C spectra were

obtained over a temperature range of 30 1C to ꢀ100 1C in

5 1C increments. Decoupled carbon spectra were recorded on

samples spun at 20 Hz using a 32 kHz sweep width and a 3 ms

pulse width corresponding to a tip angle of 451. A total of 256

transients were recorded with a 1.3 s acquisition time and a 1 s

recycle delay. The FID’s were stored as 128 k ﬁles, and

processed using standard FFT and apodizing methods using

a line broadening of 0.5 Hz throughout. Proton spectra were

recorded on samples spun at 20 Hz using a spectral width of

15 kHz and a pulse width of 5.95 ms corresponding to a 451 tip

angle. Eight scans were recorded using a 1 s acquisition time

and a 5 s recycle delay. The FID’s were stored as 32 K ﬁles and

processed using standard FFT and apodization methods

where a line broadening of 0.1 Hz was used throughout.

The half height line widths of the appropriate lines were

measured at each temperature above coalescence. Selective

inversion [spectral width (5991.6 Hz), acquisition time (1.9 s),

number of points (22 K), recycle delay (10.0 s) mixing time (37

step array ranging from 2 ms to 7 s), spin rate (0 Hz), no. scans

(16) with gradient stabilization] and T1 [ﬁrst delay (20 s), pulse

sequence: {1801, delay times (array from 0.0625 to 32 s), 901,

acqu. (1.9 s)}, spin rate (20 Hz) and no. scans (16)] experiments

Crystallography. Colorless blocks of C₃₆H₅₁P were coated

with Paratone 8277 oil (Exxon) and mounted on glass ﬁbers.

All measurements were made on a Bruker Smart CCD dif-

fractometer with graphite monochromated Mo-Ka radiation.

The data were corrected for Lorentz and polarization eﬀects

and for absorption using SADABS V2.01.⁷⁷Crystal, solution

and reﬁnement details are reported in Table 8. Selected inter-

atomic distances and angles are presented in Table 1 and full

geometrical parameters are available in Table S1, ESI.w The

solution of the structure of Dipp₃P was vexed by persistent

twinning by merohedry. Several diﬀraction data sets obtained

at room temperature failed to yield satisfactory reﬁnement.

Finally, an LT dataset (193(2) K) on a very high quality crystal

was obtained using o and j scans to a maximum y value of

26.41. The program ROTAX⁷⁸was used to determine a twin

law consisting of 1801 rotation about the [1 1 0] direct axis

direction, which allowed for solution and reﬁnement of the

structure, with twin occupancy ratio of 0.83/0.17. The non-

hydrogen atoms were reﬁned anisotropically. Hydrogen atoms

Table 8 Crystal data and structure reﬁnement of Dipp₃P

Space group

a/A

R3

16.5047(7)

c/A

g/1

Z

10.2363(9)

120

3

2.45–26.40

ꢀ20 to 20, ꢀ20 to 20, ꢀ12 to 12

5662/2205 (R_int= 0.0220)

2205/1/113

1.153

R1 = 0.0294, wR2 = 0.0884

1

y Range for data collection/1

Index ranges, hkl

Total/independent reﬂections

Data/restraints/parameters

Goodness-of-ﬁt on F²

Final R indices [I 4 2s(I)]

R indices (all data)

for H were performed at ꢀ75, ꢀ70, ꢀ65 and ꢀ60 1C. The

rates of exchange were extracted using the CIFFIT program.⁸¹

The resulting spectra were exported into MestReC⁸²and were

subsequently converted to ASCII format at a digital resolution

of 0.1 Hz point^ꢀ1. The region of interest was selected from

each spectrum and modeled using an in house program written

ꢄc

This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientiﬁque 2008

New J. Chem., 2008, 32, 214–231 | 227

View Article Online

in Matlab⁸³that simulates the spectral parameters of the half

height line width and exchange rate which is statistically

compared by superimposing the experimental spectrum on

the simulated one to give the best ﬁt.

elsewhere.¹⁷For all these FT experiments a large amplitude

(80 mV) sinusoidal waveform was superimposed onto a trian-

gular dc potential waveform which was swept at a rate of 49.68

mV s^ꢀ1. Several frequencies were employed as speciﬁed, and

the number of data points collected was 64 k. The experi-

mental data, obtained with the FT form of instrumentation,

yield current, time and applied potential as the output infor-

mation. The FT algorithm converts the time domain to the

frequency domain to give power spectra. The inverse FT

operation then separates the data into dc and harmonic

components.

Computation. The structure of Dipp₃P was modeled using

the internal geometry tools in HyperChem 7.52,⁸⁴and opti-

mized initially using the MM+ force ﬁeld. Thereafter the

semi-empirical PM3 method was used to obtain a preliminary

structure. These geometries were exported to Gaussian 98 for

Windows,⁸⁵and were then reﬁned at the HF/6-31G(d) level of

theory without symmetry constraint. From this starting point,

the structure of Dipp₃P⁺was calculated using UHF/6-31G(d)

theory. The geometry of Dipp₃P was also optimized under C₃

and D₃point-group symmetry restraints using the B3LYP/6-

31G(d) method. Geometric results are reported in Tables 1

and S1, ESI.wA detailed energy surface for the rotation-

inversion process in Dipp₃P was undertaken free of any

symmetry restrictions. The size of the problem and the avail-

able computer resources restricted us to the HF/3-21G basis

set for this extensive set of computations. The reasonableness

of the model was checked by comparison of the geometries of

the structures obtained in the ground state and in the transi-

tion state against the results of the higher-level calculations.

Two dihedral angles involving C_ipso–P–C_ipso–C_orthoatoms were

deﬁned (Scheme 1). One of these angles was driven at 101

increments, while the second was ﬁxed at 101 increments and

the remainder of the molecule was allowed to relax completely.

An energy surface was constructed in these two angular units,

following a methodology previously reported.⁶⁶

Spectroelectrochemistry. In situ electronic spectroelectro-

chemical experiments were conducted in a 10 ꢃ 1 mm i.d.

quartz cell. A coarse Pt mesh working electrode was used for

electrolysis and a Cary 5 UV-Vis-NIR spectrophotometer was

used to obtain electronic spectra under an atmosphere of N₂

gas at 20 ꢁ 2 1C. In situ EPR electrochemistry experiments

were conducted using Au micromesh electrodes⁸⁷in a conven-

tional EPR ﬂat cell on Bruker ER200D and ESP300E instru-

ments at X-band frequencies. An EG&G 273 potentiostat was

used to generate the radical cation using an applied potential

0.1 V more positive than the peak potential vs. a Cu wire

quasi-reference electrode.

EPR spectroscopy. Solutions of the analyte were prepared

ex situ in a bulk electrolysis cell as described above. In one

experiment, an aliquot of a solution prepared by the exhaus-

tive electrolysis of 10.00 mL of 1.0 ꢃ 10^ꢀ3M solution in

CH₂Cl₂(0.1 M in ⁿBu₄NPF₆) was placed in a conventional

4 mm cylindrical quartz EPR cell. The remaining solution was

made 1.0 ꢃ 10^ꢀ2M in Dipp₃P to give B10 : 1 ratio of neutral

to oxidized material. EPR spectra were recorded at 295 and at

130 K on a Bruker E380E spectrometer operating at 9.43 GHz

in CW mode. Microwave frequencies were determined with an

EIP 548A frequency counter, and g values are measured

against the F⁺line of CaO (2.0001 ꢁ 0.0002).⁸⁸Samples were

also prepared by the addition of an excess of AgClO₄, AgPF₆

and AgSbF₆to solid Dipp₃P in ‘‘T’’ cells consisting of a 4 mm

Pyrex EPR tube fused at a 751 angle to an 8 mm Pyrex reaction

tube. CH₂Cl₂was vacuum distilled onto the solid mixture, and

the tube was ﬂame sealed. Upon melting, an immediate colour

change to cherry red was observed. The EPR cell was ﬁlled

with solutions of the [Dipp₃P⁺]X^ꢀsalts in CH₂Cl₂, and the

concentration was reduced by internal distillation of the

solvent so that the spectra were close to the detection limit

of the instrument. In this way, the eﬀect of concentration

broadening on the EPR linewidths was assumed to be elimi-

nated. These spectra were obtained on a Bruker EMX 113/10

instrument also operating at X band at 18 ꢁ 2 1C. Simulations

of the EPR spectra were conducted using Bruker SimFonia

(v. 1.25)⁸⁹and PEST WinSim (v. 0.98) software.⁵²

Electrochemistry. All voltammograms were obtained at

22 ꢁ 2 1C in freshly distilled CH₃CN and CH₂Cl₂solvents

containing 0.1 M (unless otherwise indicated) ⁿBu₄NPF₆as the

supporting electrolyte and purged with dry N₂for 10 min

directly before use. In all experiments, except the EPR spectro-

electrochemistry (see below), Pt auxiliary and silver/silver

chloride reference electrodes, the latter separated from the

bulk solution by a ﬁne-porosity frit, were employed in a three-

electrode arrangement; dc measurements were undertaken

with either BioAnalytical Systems CV-50 or 100B computer-

controlled potentiostats, and simulations of the voltammo-

grams were performed using DigiSim software.⁸⁶Pt, Au and

GC macrodisk electrodes and Pt and GC microelectrodes of

known area and radii were used as working electrodes in

stationary electrode experiments. The working electrodes were

polished with 0.3 mm alumina on a clean polishing cloth

(Buehler, USA), rinsed with water, and dried with tissue

paper. For rotating disk electrode (RDE) voltammetry,

3 mm diameter GC macrodisk electrodes were used with a

Metrohm Model 628–10 RDE accessory in conjunction with a

BAS potentiostat. Bulk electrolysis was performed with the

BAS 100B potentiostat using two concentric Pt gauze basket

electrodes in a cell described elsewhere.⁴⁷Reference potential

calibration was performed against the Cp*₂Fe^+/0(Cp*₂Fe =

decamethylferrocene) and Cc^+/0(Cc = cobaltocene) redox

couples, and potentials are quoted on the Fc^+/0(Fc =

ferrocene) scale. Details of the large amplitude Fourier Trans-

form (FT) ac voltammetric instrumentation are described

Acknowledgements

We thank the Natural Sciences and Engineering Research

Council (Canada) and the Australian Research Council for

funding. An ARC Linkage-International grant (2003–2005)

facilitated this collaboration. The assistance of Si-Xuan Goh

with some of the electrochemical experiments is gratefully

ꢄc

This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientiﬁque 2008

228 | New J. Chem., 2008, 32, 214–231

View Article Online

acknowledged. G. Wolmershauser, M. Parvez and R. Mac-

donald are thanked for obtaining X-ray data sets; J. Boas and

R. Webster assisted with collection of EPR data. C. Campana

of Bruker Analytical X-ray is thanked for help with solving the

twinning problem. G. Lin and D. Iuga obtained the solid state

³¹P NMR spectra.

(c) C. Gouverd, M. Brynda, T. Berclaz and M. Geoﬀroy, J.

Organomet. Chem., 2006, 691, 72–78; (d) P. Adkine, T. Cantat, E.

Deschamps, L. Ricard, N. Mezailles, P. Le Floch and M.

Geoﬀroy, Phys. Chem. Chem. Phys., 2006, 8, 862–868; (e) T.

Sasamori, E. Mieda, N. Nagahora, K. Sato, D. Shiomi, T. Takui,

Y. Hosoi, Y. Furukawa, N. Takagi, S. Nagase and N. Tokitoh, J.

Am. Chem. Soc., 2006, 128, 12582–12588; (f) S. Sasaki, F.

Murakami and M. Yoshifuji, Organometallics, 2006, 25,

140–147; (g) S. Sasaki, F. Murakami, M. Watanabe, K. Kato,

K. Sutoh and M. Yoshifuji, J. Organomet. Chem., 2005, 690,

2664–2672; (h) S. Marque and P. Tordo, Top. Curr. Chem., 2005,

250, 43–76; (i) S. Sasaki, R. Chowhury and M. Yoshifuji, Tetra-

hedron Lett., 2004, 45, 9193–9196; (j) S. Sasaki, H. Aoki, K.

Sutoh, S. Hakiri, K. Tsuji and M. Yoshifuji, Helv. Chim. Acta,

2002, 85, 3842–3847; (k) F. Murakami, S. Sasaki and M. Yoshi-

fuji, Angew. Chem., Int. Ed., 2002, 41, 2574–2576; (l) F. Mur-

akami, S. Sasaki and M. Yoshifuji, Angew. Chem., Int. Ed., 1999,

38, 340–343; (m) K. Tsuji, S. Sasaki and M. Yoshifuji, Tetrahe-

dron Lett., 1999, 40, 3203–3206; (n) S. Sasaki, F. Murakami, M.

Murakami and M. Yoshifuji, Tetrahedron Lett., 1997, 38,

7095–7098.

8 (a) K. Sutoh, S. Sasaki and M. Yoshifuji, Inorg. Chem., 2006, 45,

992–998; (b) N. Nagahora, T. Sasamori and N. Tokitoh, Chem.

Lett., 2006, 35, 220–221; (c) C. Moser, M. Nieger and R.

Pietschnig, Organometallics, 2006, 25, 2667–2672; (d) C. Moser,

A. Orthaber, M. Nieger, F. Belaj and R. Pietschnig, Dalton

Trans., 2006, 3879–3885; (e) N. Nagahora, T. Sasamori, N.

Takeda and N. Tokitoh, Chem.–Eur. J., 2004, 10, 6146–6151;

(f) R. Pietschnig and E. Niecke, Organometallics, 1996, 15,

891–893; (g) S. Kawasaki, A. Nakamura, K. Toyota and M.

Yoshifuji, Organometallics, 2005, 24, 2983–2987.

References

1 IUPAC recommendation for R₃P is phosphane. The widespread

use of phosphine in the English speaking world leads us to use this

non-IUPAC term.

2 (a) C. A. Tolman, Chem. Rev., 1977, 77, 313–348; (b) C. A.

Tolman, J. Am. Chem. Soc., 1970, 92, 2956–2965; (c) N. G.

Andersen and B. A. Keay, Chem. Rev., 2001, 101, 997–1030; (d)

A. L. Fernandez, M. R. Wilson, A. Prock and W. P. Giering,

Organometallics, 2001, 20, 3429–3435; (e) S. Joerg, R. S. Drago

and J. Sales, Organometallics, 1998, 17, 589–599; (f) A. Fernan-

dez, C. Reyes, M. R. Wilson, D. C. Woska, A. Prock and M. P.

Giering, Organometallics, 1997, 16, 342–348.

3 (a) R. A. Stockland, M. C. Kohler, I. A. Guzei, M. E. Kastner, J.

A. Bawiec, D. C. Labaree and R. B. Hochberg, Organometallics,

2006, 25, 2475–2485; (b) V. A. Zagumennov and E. V. Nikitin,

Russ. J. Electrochem., 2003, 39, 1236–1239; (c) R. C. Bott, G. A.

Bowmaker, R. W. Buckley, P. C. Healy and M. C. S. Perera,

Aust. J. Chem., 2000, 53, 175–181; (d) A. Bayler, A. Schiera and

H. Schmidbaur, Inorg. Chem., 1998, 37, 4353–4359; (e) A. Bayler,

A. Schier, G. A. Bowmaker and H. Schmidbaur, J. Am. Chem.

Soc., 1996, 118, 7006–7007.

9 B. Laleu, G. Bernardinelli, R. Chauvin and J. Lacour, J. Org.

Chem., 2006, 71, 7412–7416, and references therein.

4 Abbreviations used in this paper. Dipp: 2,6-diisopropylphenyl.

Duryl: 2,3,5,6-tetramethylphenyl. Mes: 2,4,6-trimethyphenyl.

Mes*: 2,4,6-tri-tert-butylphenyl. Tripp: 2,4,6-triisopropylphenyl.

Xyl: 2,6-dimethylphenyl. RT: room temperature, 22 1C unless

otherwise speciﬁed. LT: low temperature. VT: variable tempera-

ture. LW: line-width. EI: electron impact. dc: direct current. ac:

alternating current. GC: glassy carbon. RDE: rotating disk

electrode. FT: Fourier transform. EPR: electron paramagnetic

resonance. SS: solid state. MAS: magic angle spinning. CP: cross-

polarization. Dn_1/2: full-width at half height. Fc: ferrocene,

Cp₂Fe. Cc: cobaltocene, Cp₂Co, Cp: Z⁵-C₅H₅. Cp*: Z⁵-

C₅(CH₃)₅. hfc: hyperﬁne coupling. LSA: line shape analysis.

CV: cyclic voltammetry. SCE: aqueous saturated calomel

electrode.

5 (a) E. C. Alyea, G. Ferguson, J. F. Gallagher and J. Malito, Acta

Crystallogr., Sect. C, 1993, 49, 1473–1476; (b) A. Bayler, G. A.

Bowmaker and H. Schmidbaur, Inorg. Chem., 1996, 35,

5959–5960; (c) E. C. Alyea, G. Ferguson and S. Kannan, Poly-

hedron, 2000, 19, 2211–2213; (d) R. C. Bott, G. A. Bowmaker, R.

W. Buckley, P. C. Healy and M. C. S. Perera, Aust. J. Chem.,

2000, 53, 175–181.

10 (a) R. T. Boere

Soc., Dalton Trans., 1998, 4147–4154; (b) R. T. Boere

and G. Wolmershauser, Can. J. Chem., 2000, 78, 583–589; (c) R.

´

, V. Klassen and G. Wolmershauser, J. Chem.

¨

´

, V. Klassen

¨

E. Boere

Chem., 2000, 78, 1613–1619; (d) R. T. Boere

Junk and J. D. Masuda, Chem. Commun., 2004, 2564–2565.

11 R. T. Boere and J. D. Masuda, Can. J. Chem., 2002, 80,

1607–1617.

´

, R. T. Boere, J. Masuda and G. Wolmershauser, Can. J.

´

¨

´

, M. L. Cole, P. C.

´

12 S. Sasaki, K. Sutoh, M. Murakami and M. Yoshifuji, J. Am.

Chem. Soc., 2002, 124, 14830–14831.

13 A recently reported system where Tripp groups are combined

with a bulky terphenyl actually has smaller sums-of-angles

around phosphorus: S. Sasaki, R. Chowdhury and M. Yoshifuji,

Tetrahedron Lett., 2004, 45, 9193–9196.

´

14 R. T. Boere and Y. Zhang, J. Organomet. Chem., 2005, 690,

2651–2657.

15 J. A. S. Howell, J. D. Lovatt, P. McArdle, D. Cunningham, E.

Maimone, H. E. Gottlieb and Z. Goldschmidt, Inorg. Chem.

Commun., 1998, 1, 118–120.

16 An analysis of six crystal structures of phosphines with three

equivalent, mono ortho substituted, aryl rings, all with substitu-

ents in the exo position, suggests that suﬃciently bulky groups

6 (a) S. Sasaki and M. Yoshifuji, Curr. Org. Chem., 2007, 11, 17–31;

(b) J. C. Lee, M. G. Wang and F. E. Hong, Eur. J. Inorg. Chem.,

2005, 5011–5017; (c) J. C. Hierso, A. Fihri and R. Amardeil,

Tetrahedron, 2005, 61, 9759–9766; (d) D. Alberti, R. Goddard and

K. R. Porschke, Organometallics, 2005, 24, 3907–3915; (e) F. E.

Hong, Y. J. Ho and Y. C. Chang, J. Organomet. Chem., 2005,

690, 1249–1257; (f) A. Yahav, I. Goldberg and A. Vigalok, Inorg.

Chem., 2005, 44, 1547–1553; (g) S. Ogoshi, M. Ueta, M. Oka and

H. Kurosawa, Chem. Commun., 2004, 2732–2733; (h) F. Bellina,

A. Carpita and R. Rossi, Synthesis (Stuttgart), 2004, 2419–2440;

(i) R. C. Smith and J. D. Protasiewicz, Organometallics, 2004, 23,

4215–4222; (j) M. Jimenez-Tenorio, M. C. Puerta, I. Salcedo, P.

Valerga, S. I. Costa, L. C. Silva and P. T. Gomes, Organome-

tallics, 2004, 23, 3139–3146; (k) A. S. King, G. Ferguson, J. F.

Britten and J. F. Valliant, Inorg. Chem., 2004, 43, 3507–3513; (l)

G. Canepa, C. D. Brandt and H. Werner, Organometallics, 2004,

23, 1140–1152; (m) H. Werner, Dalton Trans., 2003, 3829–3837;

(n) T. Matsumoto, T. Kasai and K. Tatsumi, Chem. Lett., 2002,

346–347.

P

cause a decrease in

31G(d) calculations show a smaller eﬀect from this ‘‘reverse’’

{+CPC} of up to 101; however HF/6-

steric pressure¹⁴

.

17 (a) B. D. Fleming, N. L. Barlow, J. Zhang, A. M. Bond and F. A.

Armstrong, Anal. Chem., 2006, 78, 2948–2956; (b) D. J. Gava-

ghan and A. M. Bond, Electroanalysis, 2006, 18, 333–344; (c) A.

A. Sher, A. M. Bond, D. J. Gavaghan, K. Gillow, N. W. Duﬀy, S.

X. Guo and J. Zhang, Electroanalysis, 2005, 17, 1450–1462; (d) B.

D. Fleming, J. Zhang, A. M. Bond, S. G. Bell and L. L. Wong,

Anal. Chem., 2005, 77, 3502–3510; (e) A. M. Bond, N. W. Duﬀy,

S. X. Guo, J. Zhang and D. Elton, Anal. Chem., 2005, 77,

186A–95A; (f) A. A. Sher, A. M. Bond, D. J. Gavaghan, K.

Harriman, S. W. Feldberg, M. W. Duﬀy, S. X. Guo and J. Zhang,

Anal. Chem., 2004, 76, 6214–6228; (g) J. Zhang, S. X. Guo, A. M.

Bond and F. Marken, Anal. Chem., 2004, 76, 3619–3629.

18 P. P. Power, Chem. Rev., 2003, 103, 789–809. As deﬁned herein, a

distinction is made between free radicals that are inherently stable

as isolated species, which do not decompose at RT under an inert

atmosphere, and persistent radicals which have a relatively long

lifetime under the conditions under which they are generated.

7 (a) C. Dutan, S. Shah, R. C. Smith, S. Choua, T. Berclaz, M.

Geoﬀroy and J. D. Protasiewicz, Inorg. Chem., 2003, 42,

6241–6251; (b) A. Matni, L. Boubekeur, N. Mezailles, P. Le

Floch and M. Geoﬀroy, Chem. Phys. Lett., 2005, 411, 23–27;

ꢄc

This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientiﬁque 2008

New J. Chem., 2008, 32, 214–231 | 229

View Article Online

19 A. Armstrong, T. Chivers and R. T. Boere

´

, in Modern Aspects of

38 A. D. Bain, Prog. Nucl. Magn. Reson. Spectrosc., 2003, 43,

63–103.

39 A. D. Bain and P. Hazendonk, J. Phys. Chem. A, 1997, 101,

7182–7188.

Main Group Chemistry, ed. M. Lattman and R. A. Kemp, ACS

Symp. Ser., American Chemical Society, Washington, DC, 2006,

vol. 917, pp. 66–80.

20 (a) S. L. Hinchley, C. A. Morrison, D. W. H. Rankin, C. L. B.

Macdonald, R. J. Wiacek, A. H. Cowley, M. F. Lappert, G.

Gundersen, J. A. C. Clyburne and P. P. Power, Chem. Commun.,

2000, 2045–2046; (b) S. L. Hinchley, C. A. Morrison, D. W. H.

Rankin, C. L. B. Macdonald, R. J. Wiacek, A. Voigt, A. H.

Cowley, M. F. Lappert, G. Gundersen, J. A. C. Clyburne and P.

P. Power, J. Am. Chem. Soc., 2001, 123, 9045–9053.

21 (a) S. Loss, A. Magistrato, L. Cataldo, S. Hoﬀmann, M. Geof-

froy, U. Rothlisberger and H. Grutzmacher, Angew. Chem., Int.

Ed., 2001, 40, 723–726; (b) L. Cataldo, C. Dutan, S. K. Misra, S.

Loss, H. Grutzmacher and M. Geoﬀroy, Chem.–Eur. J., 2005, 11,

3463–3468.

40 F. H. Allen, O. Kennard, D. G. Watson, L. Brammer, A. G.

Orpen and R. Taylor, J. Chem. Soc., Perkin Trans. 2, 1987,

S1–S19.

41 A. J. Bard and L. R. Faulkner, Electrochemical Methods: Funda-

mentals and Applications, Wiley, New York, 2nd edn, 2001.

42 C. M. A. Brett and M. O. Brett, Electrochemistry: Principles,

Methods and Applications, OUP, Oxford, 1993, p. 29.

43 Cited as 0.38 cP at 25 1C, in Handbook of Organic Solvent

Properties, ed. I. M. Smallwood, Wiley, NY, 1996.

44 B. Merzougui, Y. Berchadsky, P. Tordo and G. Gronchi, Elec-

trochim. Acta, 1997, 42, 2445–2453.

45 M. Matsumoto and T. W. Swaddle, Inorg. Chem., 2004, 43,

2724–2735.

22 Y. Canac, A. Baceiredo, W. W. Schoeller, D. Gigmes and G.

Bertrand, J. Am. Chem. Soc., 1997, 119, 7579–7580.

46 [Cc]^+/0ꢀ1.35 ꢁ 0.01 V vs. Fc^+/0: R. S. Stojanovic and A. M.

23 (a) A. Armstrong, T. Chivers, M. Parvez and R. T. Boere

Chem., 2004, 116, 508–511; (b) A. Armstrong, T. Chivers, M.

Parvez and R. T. Boere, Angew. Chem., Int. Ed., 2004, 43,

502–505; (c) A. Armstrong, T. Chivers, M. Parvez, G. Schatte

and R. T. Boere, Inorg. Chem., 2004, 43, 3453–3460; (d) A.

Armstrong, T. Chivers, H. M. Tuononen, M. Parvez and R. T.

Boere, Inorg. Chem., 2005, 44, 7981–7991.

´

, Angew.

Bond, Anal. Chem., 1993, 65, 56–64.

47 A. M. Bond, Broadening Electrochemical Horizons. Principles and

Illustration of Voltammetric and Related Techniques, Oxford

University Press, Oxford, 2002.

48 (a) Z. B. Alfassi, P. Neta and B. Beaver, J. Phys. Chem. A, 1997,

101, 2153–2159; (b) P. K. Wong and A. O. Allen, J. Phys. Chem.,

1970, 74, 774–778; (c) H. D. Burrows, D. Greatorex and T. J.

Kemp, J. Phys. Chem., 1972, 76, 20–26.

´

24 W. Kaim, H. Bock and H. Noth, Chem. Ber., 1978, 111,

3276–3279.

¨

49 M. Nakamura, M. Miki and T. Majima, J. Chem. Soc., Perkin

Trans. 2, 2000, 1447–1452.

25 (a) R. G. Hicks and R. Hooper, Inorg. Chem., 1999, 38, 284–286;

(b) R. G. Hicks, L. Ohrstrom and G. W. Patenaude, Inorg. Chem.,

2001, 40, 1865–1870; (c) T. Barclay, R. G. Hicks, A. S. Ichimura

and G. W. Patenaude, Can. J. Chem., 2002, 80, 1501–1506.

26 P. Agarwal, N. A. Piro, K. Meyer, P. Muller and C. C. Cummins,

¨

Angew. Chem., Int. Ed., 2007, 46, 3111–3114.

27 O. V. Parakin, E. V. Nikitin, Yu. A. Ignat’ev, A. S. Romakhin,

Yu. M. Kargin, A. V. Il’yasov, A. A. Vaﬁna, G. V. Romanov and

A. N. Pudovik, J. Gen. Chem. USSR, 1984, 55, 1058.

50 (a) M. Oyama, T. Higuchi and S. Okazaki, Electrochem. Solid-

State Lett., 2002, 5, E1–E3; (b) R. A. Kipp, J. A. Simon, M.

Beggs, H. E. Ensley and R. H. Schmehl, J. Phys. Chem. A, 1998,

102, 5659–5664; (c) T. Sumiyoshi, Chem. Lett., 1995, 645–6.

51 (a) M. P. Ferroud-Plattet, Y. Ayant, E. Belorizky and P. Tordo,

Solid State Commun., 1991, 80, 947–949; (b) C. Palau, Y. Berch-

adsky, F. Chalier, J.-P. Finet, G. Gronchi and P. Tordo, J. Phys.

Chem., 1995, 99, 158–163.

52 WinSim (v.0.98, 2002) software: D. R. Duling, J. Magn. Reson.

B., 1994, 104, 105–110.

28 R. T. Boere, J. D. Masuda and P. Tran, J. Organomet. Chem.,

´

2006, 691, 5597–5603.

29 Direct reaction of DippMgBr with PCl₃in THF produced a

complex reaction mixture, among which {DippP}₃and {DippP}₄

are major constituents (NMR and crystallographic evidence). The

synthesis of {Tripp₂P}₂is reported to form from reaction of PCl₃

with TrippLi in reasonable yield and in poor yield from

TrippMgBr: F. J. Brady, C. J. Cardin, D. J. Cardin and D. J.

Wilcock, Inorg. Chim. Acta, 2000, 298, 1–8. Two Tripp groups

can be introduced cleanly into R₂PCl using conventional lithium

reagents,^7gbut the third substitution requires an organocopper

reagent.¹²We attribute the eﬀectiveness of ArCu reagents to the

suppression of electron-transfer which induces P–P coupling

reactions.

53 The relative sizes of the isotropic ¹H–³¹P coupling constants of

the neutral phosphine were used as a guide for these simulations.

Note that the values obtained in simulation are in excellent

agreement with the calculated hfc values for these two H atom

sites, so long as dynamic averaging of the static structure is

allowed for in the solution values (average of H4 and H6 =

0.16 mT; of H7 and H10 = 0.10 mT)—see Table S3, ESIw.

54 (a) A. V. Il’yasov, Yu. M. Kargin, E. V. Nikitin, A. A. Vaﬁna, G.

V. Romanov, O. V. Parakin, A. A. Kazakova and A. N. Pudovik,

Phosphorus Sulfur, 1980, 8, 259–262; (b) A. V. Il’yasov, Yu. M.

Kargin and A. A. Vaﬁna, Russ. J. Gen. Chem., 1993, 63,

1833–1840.

30 A. Aranyos, D. W. Old, A. Kiyomori, J. P. Wolfe, J. P.

Sadighi and S. L. Buchwald, J. Am. Chem. Soc., 1999, 121,

4369–4378.

31 S. J. Greenﬁeld and S. R. Gilbertson, Synthesis (Stuttgart), 2001,

2337–2340.

55 N. G. Connelly and W. E. Geiger, Chem. Rev., 1996, 96, 877–910.

56 L. Song and W. C. Trogler, Angew. Chem., Int. Ed. Engl., 1992,

31, 770–772. The review by Connelly and Geiger lists the potential

of the Ag^0/+couple in CH₃CN solution as only +0.04 V vs.

Fc^+/0. Consistent with this value, we found that AgClO₄in

CH₃CN is incapable of generating the Dipp₃P⁺radical.

ꢂ

32 F. Langer, K. Punterer, R. Sturmer and P. Knochel, Tetrahedron:

¨

Asymmetry, 1997, 8, 715–738.

¨

57 (a) J. F. Blount, C. A. Maryanoﬀ and K. Mislow, Tetrahedron

Lett., 1975, 913–916; (b) P. Finocchiaro, D. Gust and K. Mislow,

J. Am. Chem. Soc., 1974, 96, 2165–2167; (c) J. D. Andose and

K. Mislow, J. Am. Chem. Soc., 1974, 96, 2168–2176; (d) K.

Mislow, Acc. Chem. Res., 1976, 9, 26–33.

58 N. N. Greenwood and A. Earnshaw, Chemistry of the Elements,

Butterworth-Heinemann, Oxford, 2nd edn, 1997, p. 493.

59 (a) D. G. Gilheany, Chem. Rev., 1994, 94, 1339–1374; (b) T. A.

Albright, J. K. Burdett and M.-H. Whangbo, Orbital Interactions

in Chemistry, Wiley, New York, 1985.

33 H.-F. Grutzmacher and D. Kirchhoﬀ, Organometallics, 2001, 20,

¨

3738–3744.

34 (a) M. Nespolo and G. Ferraris, Acta Crystallogr., Sect. A, 2004,

60, 89–95; (b) M. Nespolo and G. Ferraris, Z. Kristallogr., 2003,

218, 178–181.

35 (a) P. Changenet, P. Plaza, M. M. Martin, Y. H. Meyer and W.

Rettig, Chem. Phys., 1997, 221, 311–322; (b) N. A. Rozanel’skaya,

A. I. Bokanov, B. M. Uzhinov and B. I. Stepanov, J. Gen. Chem.

USSR, 1975, 45, 277–280.

36 (a) M. Culcasi, Y. Berchadsky, G. Gronchi and P. Tordo, J. Org.

Chem., 1991, 56, 3537–3542; (b) J. F. Blount, D. Camp, R. D.

Hart, P. C. Healy, B. W. Skelton and A. H. White, Aust. J.

Chem., 1994, 47, 1631–1639; (c) A. I. Bokanov, P. Yu. Ivanov, N.

A. Rozanel’skaya and B. I. Stepanov, Russ. J. Gen. Chem., 1975,

47, 702–705.

37 (a) P. S. Pregosin and R. Kunz, NMR: Basic Principles and

Progress, 1979, 16, 47–55; (b) M. J. T. Ditty and W. P. Power,

Can. J. Chem., 1999, 77, 1951–1961, and references therein.

60 It is well-known that longer E–Si bonds allow for much lower

barriers to pyramidal inversion, and that the ground state struc-

tures of trisilylamines are often planar.⁵⁹By contrast, P(SiR₃)₃

usually retain pyramidal ground state structures. For example, in

Mes*P(SiMe₃)₂, +(C, SiPC, Si) = 343.2(1)1 (a) A. H. Cowley,

M. Pakulski and N. C. Norman, Polyhedron, 1987, 6, 915–919. In

P

P(SiMe₃)₃, +(SiPSi) = 318.071; (b) J. Bruckmann and C.

Kruger, Acta Crystallogr., Sect. C, 1995, 51, 1152–1155. How-

P

ever, in the more congested P(SiⁱPr)₃, +(SiPSi) = 359.789(3)1;

ꢄc

This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientiﬁque 2008

230 | New J. Chem., 2008, 32, 214–231

View Article Online

(c) M. Driess, K. Merz and C. Monse, Z. Anorg. Allg. Chem.,

2000, 626, 2264–2268; (d) C. von Hanisch, Z. Anorg. Allg. Chem.,

Electrochim. Acta, 1994, 39, 2395–2400; (c) P. Tordo, in Landolt-

Bornstein, New Series, Magnetic Properties of Free Radicals, ed.

H. Fischer, Springer, Berlin, 1988, Group 2, vol. 17e, Part 9, pp.

254–313.

¨

2001, 627, 1414–1416. The implication of these results is that P–Si

conjugation signiﬁcantly lowers the barrier to pyramidal inver-

sion compared to P–C bonds, but that the eﬀect is smaller than

for N; though pyramidal remains the preferred geometry, steric

eﬀects compete eﬀectively.

73 V. A. Zagumennov and E. V. Nikitin, Russ. J. Electrochem., 2003,

39, 1236–1239.

74 S. Yamaguchi, S. Akiyama and K. Tamao, J. Organomet. Chem.,

2002, 646, 277–281.

75 R. R. Schrock, M. Wesolek, A. H. Liu, K. C. Wallance and J. C.

Dewan, Inorg. Chem., 1988, 27, 2050–2054.

76 An alternative route for the synthesis of Dipp₃P is the reaction of

puriﬁed DippPCl₂with two equivalents of (DippCu)_xin aliphatic

hydrocarbons (e.g. reﬂuxing n-heptane). We found that the yields

from this procedure are considerably lower than that of the in situ

procedure described here, despite the fact that the copper reagent

has been structurally characterized and is prepared in pure

form.²⁸Solvent may play a role.

77 Bruker AXS Inc., Bruker SMART v 5.631, 2003.

78 ROTAX: Simon Parsons & Bob Gould, University of Edinburgh

with additions by Richard Cooper (Oxford) and Louis Farrugia

(Glasgow). Version 26th November, 2001.

61 G. H. Penner and R. E. Wasylishen, Can. J. Chem., 1989, 67,

1909–1913.

62 C. A. Jolly, F. Chan and D. S. Marynick, Chem. Phys. Lett., 1990,

174, 320–324.

63 A. Rieker and H. Kessler, Tetrahedron Lett., 1969, 1227–1230.

64 (a) B. I. Stepanov, V. M. Matyuk, A. I. Bokanov and E. N.

Karpova, J. Gen. Chem. USSR, 1975, 45, 2059–2060; (b) V. V.

Negrebetskii, A. I. Bokanov, L. Ya. Bogel’fer, N. A. Rozanel’s-

kaya and B. I. Stepanov, J. Gen. Chem. USSR, 1978, 48,

1196–1197; (c) V. V. Negrebetskii, A. I. Bokanov, N. A. Roza-

nel’skaya and B. I. Stepanov, J. Gen. Chem. USSR, 1979, 49,

1304–1306; (d) V. V. Negrebetskii, A. I. Bokanov, L. Ya. Bogel’-

fer, N. A. Rozanel’skaya and B. I. Stepanov, J. Gen. Chem.

USSR, 1978, 48, 1196–1200.

65 A. J. Bellamy, R. O. Gould and M. D. Walkinshaw, J. Chem.

Soc., Perkin Trans. 2, 1981, 1099–1104.

79 G. M. Sheldrick, SHELXTL. Version 6.14, Bruker AXS Inc.,

Madison, WI, USA, 2003.

80 VNMR manual. Varian, Inc., USA, 2000.

66 H. E. Birkett, J. C. Cherryman, A. M. Chippendale, P. Hazen-

donk and R. K. Harris, J. Mol. Struct., 2002, 602–603, 59–70.

67 (a) A. S. Romakhin, E. V. Nikitin, O. V. Parakin, Yu. A.

Ignat’ev, B. S. Mironov and Yu. M. Kargin, J. Gen. Chem.

USSR, 1985, 56, 2298–2301; (b) A. S. Romakhin, F. M. Palyutin,

Yu. A. Ignat’ev, E. V. Nikitin and Yu. M. Kargin, J. Gen. Chem.

USSR, 1987, 58, 1723–1724; (c) A. S. Romakhin, V. A. Zagu-

mennov, E. V. Nikitin and Yu. M. Kargin, Russ. J. Gen. Chem.,

1995, 65, 1208–1212; (d) F. Chalier, Y. Berchadsky, J.-P. Finet,

G. Gronchi, S. Marque and P. Tordo, J. Phys. Chem., 1996, 100,

4323–4330; (e) B. Merzougi, Y. Berchadsky, P. Tordo and G.

Gronchi, Electrochim. Acta, 1997, 242, 2445–2453.

68 G. Gritzner and J. Kuta, Pure Appl. Chem., 1984, 56, 461–466.

69 Interestingly, the contribution of methyl groups to the redox

potentials of metallocenes have similar magnitudes: Cp*₂Fe^+/0

is about 0.33 V more cathodic than Fc^+/0, equivalent to about

0.03 V per methyl group.

81 A.D. Bain, 1995; based on the SIFFIT program described in: D.

R. Muhandiram and R. E. D. McClung, J. Magn. Reson., 1987,

71, 187–192.

82 C. Cobas, J. Craces and J. Sandina, MestRe-C 2.3a, Dept. de

Quimica Organica, Univ. de Santiago de Campostella, Spain,

1996–2000.

83 Mathworks, Inc., Matlab 6.5, 2002.

84 Hypercube, Inc., HyperChem Release 7.52 for Windows.

85 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,

M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A.

Montgomery, Jr., R. E. Stratmann, J. C. Burant, S. Dapprich,

J. M. Millam, A. D. Daniels, K. N. Kudin, M. C. Strain,

O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi,

B. Mennucci, C. Pomelli, C. Adamo, S. Cliﬀord, J. Ochterski,

G. A. Petersson, P. Y. Ayala, Q. Cui, K. Morokuma,

D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman,

J. Cioslowski, J. V. Ortiz, A. G. Baboul, B. B. Stefanov,

G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. Gomperts,

R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y.

Peng, A. Nanayakkara, C. Gonzalez, M. Challacombe,

P. M. W. Gill, B. G. Johnson, W. Chen, M. W. Wong,

J. L. Andres, M. Head-Gordon, E. S. Replogle and J. A. Pople,

GAUSSIAN 98 (Revision A.9), Gaussian, Inc., Pittsburgh, PA,

1998.

70 The ‘‘pyramidalization angle’’ is deﬁned in refs. 36a and 51b as

the angle by which the P–C bond is bent down from the

horizontal (i.e. ﬂat geometry.) This is not easily measured from

modern software that relies on internal coordinates, or from

compiled X-ray crystallographic data, for which the sums of

angles around phosphorus +(CPC) is much easier to obtain.

P

Tordo’s a can be obtained from such data using the following

P

trigonometric relationship: a = cos^ꢀ1[sin{ +(CPC)/6}/cos 301].

71 While it is tempting to apply similar reasoning in the comparison

with unsubstituted Ph₃P and both Xyl₃P and Mes₃P, comparisons

with fully irreversible and reversible redox couples must be

discouraged; the eﬀect of follow-up reactions on the potential

86 DigiSim for Windows 95, Version 3.05; Bioanalytical Systems Inc.,

West Lafayette, IN, 2000.

87 A. Neudeck and L. Kress, J. Electroanal. Chem., 1997, 437,

141–156.

88 J. E. Wertz, P. Auzins and J. W. Orton, Discuss. Faraday Soc.,

1961, 31, 140–150.

of Ph₃P^0/+are in fact likely to have shifted the apparent E1 in the

cathodic direction.

72 (a) G. Schiavon, S. Zercchin and G. Cogoni, J. Electroanal.

Chem., 1973, 48, 425–431; (b) J. A. Caram and E. J. Vasini,

89 Bruker Analytische Messtechnik GmbH, WINEPR SimFonia

version 1.25, 1996.

ꢄc

This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientiﬁque 2008

New J. Chem., 2008, 32, 214–231 | 231

Article Doi

Photophysical, dynamic and redox behavior of tris(2,6-diisopropylphenyl) phosphine

DOI: 10.1039/b709602j

Source and publish data:

Authors:

Article abstract of DOI:10.1039/b709602j

Full text of DOI:10.1039/b709602j

Products guided by the article

R&D Labs maybe for 856259-64-8

Relevant to this article

Hot Product