Proton sponge phosphanes: Reversibly chargeable ligands for ESI-MS analysis

Article abstract of DOI:10.1002/ejic.201100820

Proton sponge phosphanes are unusual ligands in that they possess two basic sites with very different functions: the phosphorus binds to a metal center, while the 1,8-bis(dimethylamino)naphthalene binds protons strongly. Both groups are selective for thei

Full text of DOI:10.1002/ejic.201100820

FULL PAPER
DOI: 10.1002/ejic.201100820
Proton Sponge Phosphanes: Reversibly Chargeable Ligands for ESI-MS
Analysis
Nicola J. Farrer,[a][‡] Krista L. Vikse,[a] Robert McDonald,[b] and J. Scott McIndoe*[a]
Keywords: Proton sponge / Phosphanes / Mass spectrometry / Organocatalysis
Proton sponge phosphanes are unusual ligands in that they
possess two basic sites with very different functions: the
phosphorus binds to a metal center, while the 1,8-bis(dimeth-
ylamino)naphthalene binds protons strongly. Both groups are
selective for their target, and the binding of the proton is
easily reversible by controlling the pH of the solution. Proton
sponge phosphanes therefore make useful chargeable li-
gands for the purposes of electrospray ionization mass spec-
trometric analysis of the complexes in which they are bound.
However, where the 1,8-bis(dimethylamino)naphthalene is
functionalized with the phosphane is critical, and the 1-di-
methylamino group can become involved in binding to the
metal when diphenylphosphane is in the 2-position. This be-
havior is undesirable from the point of view of keeping pro-
ton and metal binding segregated, and we introduce an ef-
ficient synthesis of the 4-substituted proton sponge phos-
phane in which the two functional groups are on opposite
sides of the naphthalene ring. This precaution solves the
problem of the molecule acting as a bidentate ligand, but
caution is still needed when using the ligand, because ioniza-
tion pathways such as halide dissociation to form [M – X]+
ions and oxidation of electron-rich metal complexes to form
[M]·+ can be competitive with protonation.
studies of various palladium and platinum organometallic
complexes.[2] Since then, a growing application of ESI-MS
in this area has been in the identification of short-lived,
low-concentration intermediates. It has been used in the
Introduction
Electrospray ionization mass spectrometry (ESI-MS) has
a number of strengths that make it well suited to the study
of organometallic catalysis. (1) ESI-MS is a soft technique
that operates on solutions and can leave weak bonding in-
teractions intact. (2) Only species that are already charged
in solution or contain an easily charged site are detected.
Because of this most common solvents are “invisible” and
very low detection limits are accessible.[20] (3) Analysis is
fast (on the order of seconds), and (4) intermediates at nano-
molar concentrations can be detected with ease. Finally,
(5) since each species in solution is usually represented by
a single peak in the mass spectrum it is often possible to
extract information from complex mixtures. A growing
body of literature exists in which investigators have taken
advantage of these attributes of ESI-MS to study organo-
metallic systems. The first was Berman who used ESI-MS
to detect a number of environmentally important organo-
arsenic ions.[1] Another notable early example comes from
Canty in 1993 who reported the positive ESI-MS(/MS)
study of catalytic oxidation,[3] hydrogenation,[4] hydro-
[5]
silylation
and carbon–carbon bond-forming reactions.[6]
A large amount of the attention has been given to palla-
dium-catalyzed carbon–carbon bond-forming reactions.[7]
In order to investigate any catalytic system by ESI-MS all
species of interest must be charged; either inherently, adven-
titiously (e.g., by protonation or loss of a halide), or inten-
tionally by installing a charged or chargeable tag.
Monitoring catalytic reactions that have intrinsically or
adventitiously charged intermediates is relatively simple,
and analyses of these types of systems constitute the bulk
of the literature,[8] but many of the most important catalytic
organometallic reactions proceed through neutral interme-
diates where there are no reliable ionization mechanisms for
visualization by ESI-MS. In order to study these systems a
charged or chargeable (usually having an acidic or basic
site) tag is required. Importantly, the tag must not introduce
steric or electronic effects that interfere with the catalysis in
any way.
In 1994 Canary et al. purposefully used a substrate with
an easily protonated site to study the palladium-catalyzed
reactions of pyridyl bromide with three different phenyl-
boronic acids by ESI-MS. Pyridyl bromide was selected as
a chargeable tag due to the ability of the ring nitrogen to
become protonated. Relying on this ionization mechanism,
oxidative addition intermediates and transmetallation inter-
[a] Department of Chemistry, University of Victoria,
P. O. Box 3065, Victoria, BC V8W 3V6, Canada
E-mail: mcindoe@uvic.ca
[b] X-ray Crystallography Laboratory, Department of Chemistry,
University of Alberta,
Edmonton, AB, T6G 2G2, Canada
[‡] Current address: Department of Chemistry, University of
Warwick,
Coventry CV4 7AL, UK
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejic.201100820.
Eur. J. Inorg. Chem. 2012, 733–740
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
733

N. J. Farrer, K. L. Vikse, R. McDonald, J. S. McIndoe
FULL PAPER
mediates were observed.[9] Many of the existing studies bind to 1, so mass spectra of 1 display only the [M + H]+
using charged ligands focus on ruthenium-catalyzed sys- peak even in the presence of competing cations. Its high
tems, and they often make use of ligands designed to confer ionization efficiency in solution leads to signal intensities
water solubility. The groups of Traeger,[10] Nicholson,[11] approaching those of phosphonium ions allowing for detec-
and Chen[12] have made significant contributions to this tion at very low concentrations and ensuring that all species
area. However, it is still a somewhat under-appreciated ap- containing 1 are represented in the spectrum. Di-proton-
proach.
ation is extremely unfavorable and the doubly charged ion
Chargeable tags are neutral molecules that contain a is never seen, further simplifying the resulting spectra. Ap-
functional group of interest, and at a secondary site bind pending 1 to common ligands was anticipated to provide a
either an acid or a base to become charged. Chargeable tags means of reversibly and selectively charging and investiga-
(as opposed to permanently charged tags) are appealing for ting a variety of organometallic and coordination com-
a number of reasons: (i) Any effect that the charged group pounds and their reactivity. Triphenylphosphane was se-
may have on reactivity is mitigated since the charged tag is lected as the parent ligand since it is inexpensive, readily
in equilibrium with its neutral counterpart, (ii) the pH of available and is used routinely as a ligand in many catalytic
the droplet decreases rapidly as the solvent evaporates,[13] reactions. 1 can be substituted for one of the phenyl groups
so addition of protons is often not required, and (iii) on phosphorus to give the ortho- or para-electrospray-active
chargeable ligands have the potential to aid in catalyst re- ligands
1,8-bis(dimethylamino)-2-(diphenylphosphanyl)-
covery since their solubility in organic solvents can be ma- naphthalene (3a) or 1,8-bis(dimethylamino)-4-(diphenyl-
nipulated by changing the pH of the reaction mixture.
We have developed chargeable phosphane ligands de-
rived from the strong base 1,8-bis(dimethylamino)naphth-
alene also known as “Proton Sponge®” that serve as ana-
logues for commonly used organometallic ligands.[14] First
studied by Alder and co-workers, Proton Sponge® (1) is a
very strong, non-nucleophilic base (pKa = 12.34 in H2O)
with high thermodynamic basicity but relatively low kinetic
basicity.[15] 1 binds H+ by forming intramolecular hydrogen
bonds with both nitrogen centers in the molecule to give
the charged [N···H···N]+ moiety. 1 has been used extensively
as a non-nucleophilic proton abstractor in organic synthe-
sis,[16] and as a matrix material in MALDI-MS.[17]
phosphanyl)naphthalene (3b) (see Scheme 1).
The high thermodynamic basicity of 1 is attributed to a
combination of factors.[18] The methyl groups bound to
each nitrogen sterically clash with each other and force the
nitrogen lone pairs to point toward each other adopting an
“in-in” conformation.[19] This leads to a high degree of lone
Scheme 1. Synthesis of 3a and 3b. a) NBS / THF / –78 °C; 63%
yield. b) Br2 / CCl4 / 22 °C; 52% yield. c) nBuLi / THF / –78 °C,
PPh2Cl / THF / –78 °C.
Initial work on the two isomers revealed that while syn-
pair-lone pair repulsion and causes the naphthalene back- thesis of 3a was straightforward, synthesis of 3b was not.
bone to be twisted. By protonating the neutral molecule a The problem lies in the first step of the reaction: bromina-
more optimal geometry is attained, steric strain is relieved, tion of 1, which can be effectively accomplished in the ortho
and the naphthalene ring returns to planarity.[20] Additional (2-) position using the brominating agent N-bromosuc-
stabilization is gained through the strong intramolecular hy- cinimide, selective for ortho-bromination at low tempera-
drogen bonds formed between the proton and both nitrogen tures.[22] However, despite attempting the reaction under a
lone pairs.[18a] The low kinetic basicity of 1 arises from the variety of brominating conditions that were reported in the
steric crowding of the basic site by the four nearby methyl literature,[23] selective para-bromination at only one of the
groups. These methyl groups and the proximity of the two para-sites on Proton Sponge® was elusive. Product mixtures
nitrogen centers to each other are also responsible for the which included ortho-substituted and di-substituted com-
high selectivity that 1 has for H+; all other cations are too pounds were obtained, and purification of 3b from these
large to fit in the resulting pocket.[21] The overall effect is mixtures was difficult and low yielding.
that 1 is approximately six orders of magnitude more basic
3a reacts with a range of metal complexes, including
than typical aromatic amines, and can selectively trap ad- Fe(CO)5, W(CO)5(thf), PdCl2(cod) (cod = cyclooctadiene),
ventitious protons and become charged.
and {Ru(η6-p-cymene)Cl2}2. In the reactions with Fe(CO)5
and W(CO)5(thf), 3a coordinated only through the phos-
phorus atom, and the ESI-MS showed only the expected
[M + H]+ ions.[14] Reaction of 3a with [Ru(η6-p-cymene)-
Cl2]2 produced Ru(η6-p-cymene)Cl2(3a) as a crystalline
Results and Discussion
We expected 1 to be a well-behaved electrospray-active product, but the small orange prisms obtained did not give
tag based on the properties discussed above. Other poten- enough reflections for an X-ray crystallographic structure
tially interfering cations such as Na+ or K+ are too large to to be determined. The ESI-MS (in different solvents, in-
734
www.eurjic.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2012, 733–740

Proton Sponge Phosphanes
Figure 1. Positive-ion ESI-MS and MS/MS of [Ru(η6-p-cymene)Cl(3a)]+ in CH2Cl2. Fragmentation by loss of p-cymene rather than loss
of 3a is strong evidence that 3a is strongly bound and hence not monodentate.
cluding CH2Cl2, MeOH, and MeCN) suggested that a
chloride ligand had been displaced, as the product appeared
to be [Ru(η6-p-cymene)Cl(3a)]+ rather than [Ru(η6-p-
cymene)Cl2(3a) + H]+. Assessment of NOE interactions
(2D 1H NOESY) allowed assignment of the 1H NMR spec-
trum (see Supporting Information Figure SI2) to give the
structure shown in Figure 1. The bulky proton sponge
group is seen under the methyl end of the p-cymene, rather
than the isopropyl end, which is attributed to steric interac-
tions. The protons on the dimethylamino groups nearest to
the ruthenium center are deshielded (4.18, 4.02 ppm) and
chemically distinct from each other [by way of comparison,
in the Fe complex Fe(CO)4(3a) the same protons are seen
as one signal at δ = 2.41 ppm], indicative of a substantial
nitrogen lone pair interaction with the ruthenium.
Figure 2. Single-crystal X-ray structure of trans-PdCl2(3a)2. Key
Evidently the proton sponge functionality was coordinat-
ing to the metal in this case and so further complexes were
investigated to determine the generality of this behavior.
Reaction of PdCl2(cod) with 3a produced PdCl2(3a)2
through displacement of the weakly bound cod ligand by
two equivalents of 3a, and an X-ray crystal structure deter-
mination was carried out (Figure 2, CCDC-837948).
bond lengths: Pd–Cl 2.312 Å; Pd–Cl 2.349 Å; P–C31 1.824 Å; C42–
N1 1.427 Å; Me–N (1.45Ϯ0.015) Å. Key bond angles: P–Pd–Cl
(90Ϯ1)°, Pd–P–C31 113.99°, Pd–P–C21 103.12°, Pd–P–C31
125.87°. Key torsion angle: N1–C42–C34–N2 9.00°.
ESI-MS of trans-PdCl2(3a)2 in CH2Cl2 or MeOH gave
no discernible signal, but mixtures of the solvents or ad-
The structure shows an essentially conventional triaryl- dition of a small amount of formic acid enabled detection
phosphane ligand, with some distortion due to the bulk of of charged derivatives of the complex (Figure 3). Free li-
the proton sponge in the ortho position. The considerable gand [3a + H]+ at m/z 399.06 was the most intense signal,
bulk of 3a is evident from the structure, so it is therefore less intense (palladium) species were seen at higher values.
not surprising that the trans isomer is favored. The phos- The parent complex PdCl2(3a)2 ionized principally through
phane ligand in trans-PdCl2(3a)2 is calculated to have a two different pathways: loss of Cl– gave [PdCl(3a)2]+ at m/z
cone angle[24] of 169° with measurements taken directly 938.91 whereas addition of H+ gave [PdCl2(3a)2 + H]+ at
from the crystal structure. The cone angle of PPh3 in the m/z 974.89.[26] In the presence of formic acid, the intensity
triphenylphosphane analogue trans-PdCl2(PPh3)2 is 146° of these two peaks was approximately equal despite the
when calculated by the same method.[25]
large excess of acid available. In CH2Cl2/MeOH with no
Eur. J. Inorg. Chem. 2012, 733–740
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
735

N. J. Farrer, K. L. Vikse, R. McDonald, J. S. McIndoe
FULL PAPER
Figure 3. Top: ESI-MS (CH2Cl2/formic acid) of PdCl2(3a)2. Middle: enlargement of m/z 920–1000 region. Bottom: MS/MS of [PdCl2-
(3a)2 + H]+. MS/MS of the species at m/z 485.95 and m/z 581.89 are shown in the Supporting Information.
added acid the loss of Cl– was dominant. The intensity of the metal. We therefore examined a palladium complex that
the palladium species in the presence of acid was much did not contain chloride, namely Pd(dba)2 (dba = dibenz-
lower than in CH2Cl2/MeOH, as the acid causes some de- ylideneacetone), which is a useful precursor to Pd0 phos-
composition of the original complex to generate a small phane complexes, as the dba ligands are easily displaced.
quantity of free ligand [3a + H]+. [PdCl(3a)2]+ showed a Unfortunately, complicated reactivity ensued and no species
strong affinity for CH3CN despite only trace amounts being of the form [Pd(3a)n + H]+ (n = 1, 2, 3) were detected.
present in the mass spectrometer; [PdCl(3a)(CH3CN)]+ is Instead, the Pd0 appeared to activate the proton sponge
seen at m/z 581.89, and the unsolvated complex [PdCl- functional group and species of the form [1 – H]+ were de-
(3a)]+ is seen at a lower intensity at m/z 538.90. Fragmenta- tected, suggesting the ligand had undergone C–P bond acti-
tion of [PdCl(3a)(CH3CN)]+ showed facile loss of the vation. The final nail in the coffin for 3a was that it did not
CH3CN at low collision energies (Supporting Information, promote cross-coupling reactions, including simple Suzuki
Figure SI2). The acetone adduct [PdCl(3a)(Me2CO)]+ was and Sonogashira couplings that worked well with tri-
detected at m/z 614.77. The complexity of the palladium phenylphosphane. 3a was therefore abandoned as a useful
speciation led us to synthesize the platinum equivalent, ligand, as the proximity of the proton sponge functionality
PtCl2(3a)2, which did ionize predominantly by addition of to the metal center compromised the reactivity we hoped to
a proton rather than by chloride loss (see Supporting Infor- study: it failed the test of a useful chargeable tag for ESI-
mation, Figure SI3), though the high abundance of free [3a MS studies.
+ H]+ confirmed that 3a was rather labile.
This nonideal reactivity caused us to revisit 1, substituted
The propensity of the ligand to ionize through chloride in the para position as a charged ligand with the aim of
displacement was troubling, because ideally the chargeable inhibiting unwanted participation of the proton sponge
group should provide a unique ionization pathway that functionality in reactions since the chargeable dimeth-
does not involve alteration of the coordination sphere of ylamino groups of 3b are further removed than those of 3a
736
www.eurjic.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2012, 733–740

Proton Sponge Phosphanes
and are unable to interfere with reactivity at the metal cen- PdCl2(cod) in dichloromethane was combined with a solu-
ter. A synthesis was developed under which 3b could be tion of 3b in dichloromethane and an ESI-MS of the mix-
produced with an overall isolated yield of 19% in two steps ture was collected. The major signal in the spectrum (m/z
(Scheme 1). 2b was synthesized by dropwise addition of 975) corresponds to [PdCl2(3b)2 + H]+ in which one of the
bromine (diluted in carbon tetrachloride) to a carbon tetra- chargeable ligands is protonated. In contrast to 3a, no sig-
chloride solution of 1 at room temperature. This addition nal corresponding to ionization by loss of chloride (m/z
must be performed slowly (ca. 2.5 h) to maintain selectivity 940) was observed, and very little of the free ligand is ob-
for 3b. After addition of aqueous solutions of sodium thio- served (m/z 399) suggesting that 3b performs well as a li-
sulfate and sodium hydroxide, ESI-MS showed good selec- gand for palladium.
tivity for addition of only one bromine atom to 1 (only
small amounts of 1 at m/z 215, and the di-brominated side
product, m/z 457, are observed); although MS could not
distinguish between ortho- and para-brominated isomers
(Supporting Information, Figure SI4). After workup, 1H
NMR confirmed that the correct isomer was formed (Sup-
porting Information, Figure SI5) in a ratio of 93:7 (para:or-
tho), the reverse of the ratio obtained in the synthesis of 2a
when NBS at –78 °C is used for bromination (9:91). 2b was
used without further purification to form 3b via lithiation
and treatment with PPh2Cl at low temperature in THF. The
mixture was warmed to room temperature with stirring and
then allowed to settle. After filtration, deep orange crystals
of 3b were formed in the filtrate via solvent evaporation.
An X-ray crystal structure determination was performed on
a single crystal of 3b (Figure 4, CCDC-837947).
Figure 5. Positive-ion ESI-MS of a dichloromethane solution of
PdCl2(cod) and 3b. Small peaks are observed at: m/z 415 for the
oxidized ligand [3bO + H]+, m/z 488 for the doubly charged com-
plex [PdCl2(3b)2 + 2H]2+, and m/z 539 and m/z 1077 for the singly
and doubly protonated versions of the dimer Pd2Cl2(3b)2, respec-
tively.
Given the success of 3b as a chargeable tag, we next em-
ployed 3b in the study of a catalytic reaction. Our goal was
to study palladium-catalyzed systems, and to that end we
chose a Stille reaction, which Santos had previously studied
Figure 4. Single-crystal X-ray structure of 3b. Key bond lengths:
P1–C (1.838Ϯ0.002) Å; C1–N1 1.402 Å; C9–N2 1.404 Å; N–CMe
by ESI-MS,[28] to test the performance of our chargeable
(1.452Ϯ0.002) Å. Key bond angles: C21–P1–C31 101.98°; C21–
P1–C4 100.94°; C31–P1–C4 102.28°. Key torsion angle: N1–C1–
C9–N2 22.75°.
ligand. An acetonitrile solution containing the catalyst
Pd(OAc)2 and 3b was examined by MS, and the result was
somewhat disappointing (Figure 6). While species consis-
tent with the ones reported by Santos were detected, more
Comparison of the structure of 3b to that of 3a[14] reveals than one ionization pathway was operative; protonation of
that the torsion angle reflecting the in-and-out of plane 3b was observed, and many of the palladium-containing
bending of the dimethylamino groups is 22.8° in 3b, signifi- species were also oxidized to form radical cations. The latter
cantly less than the 37.4° seen for 3a. The reduced strain is ionization method involves oxidation of Pd0 at the tip of
no doubt due to the –PPh2 group being moved from the the electrospray capillary – the very method that Santos
proximal 2- to the remote 4-position. The N···N distance is relied on in his experiments to obtain positive ions. We had
correspondingly reduced, from 2.90 to 2.78 Å. Like 3a, 3b anticipated that protonation of 3b would be correspond-
displays pH-dependent water solubility[27] – the unproton- ingly more facile, suppressing oxidation of palladium. How-
ated form is soluble in organic solvents only, addition of ever, this was not the case.
acid protonates the sponge functionality and allows aque-
ous dissolution.
Why do the radical cations form? ESI-MS owes much of
its sensitivity to the creation of an excess of positive (or
The mass spectrum of 3b gave the expected single, intense negative, depending on the ionization mode) charge on the
[M + H]+ peak (Figure 5). To test the behavior of the tag droplets. It does so through electrochemistry – a tiny
when coordinated to a metal center, a solution of amount of material is oxidized at the capillary. Typically,
Eur. J. Inorg. Chem. 2012, 733–740
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
737

N. J. Farrer, K. L. Vikse, R. McDonald, J. S. McIndoe
FULL PAPER
ization by protons, even in the presence of other cations.
Their ionization efficiency is good. They coordinate readily
to metal complexes through the phosphorus. However, they
have weaknesses that hamper their ability to perform reli-
ably: their ionization is dependent on the acidity of the
solution, they are bulkier than the parent triphenylphos-
phane (especially for 3a, the ortho-substituted sponge), co-
ordination by the sponge functionality can be problematic,
and reactivity can be substantially different to the parent
triphenylphosphane.
Experimental Section
General: Dry solvents were obtained from an MBraun solvent puri-
fication system. All solvents were HPLC grade unless otherwise
stated. Reagents were purchased from Aldrich and used without
further purification except for 1, which was recrystallized from hot
methanol. Reactions performed under nitrogen were carried out
using standard Schlenk techniques. All electrospray ionization
mass spectra were collected with a Miromass Q-TOF micro instru-
ment. Capillary voltage was set to 2900 V, source temperature was
80 °C and desolvation temperature was 150 °C. Samples were in-
fused via syringe pump at 10 μLmin–1. NMR spectra were re-
corded with a Bruker AC-300 spectrometer. Chemical shifts are
quoted in ppm using internal references of CDCl3 (1H δ =
7.26 ppm) or CD3CN (1H δ = 1.94 ppm) where appropriate. Melt-
ing points were recorded with a Gallenkamp melting point appara-
tus and are uncorrected. IR spectra were recorded using a solution
cell in a Perkin–Elmer Spectrum 1000 FT-IR spectrometer.
Figure 6. Positive-ion ESI-MS of an acetonitrile solution of
Pd(OAc)2 and 3b.
this process is invisible because the oxidized material, often
the iron in the stainless steel capillary, rarely appears in the
mass spectrum (its surface activity is low). However, Pd0
phosphane complexes are both greasy (hydrophobic, sur-
face active) and very easily oxidized, more so than the capil-
lary.
Further investigation of the Stille reaction using 3b pro-
duced spectra containing many peaks that were difficult to
assign (Supporting Information, Figure SI6). We attributed
these peaks to the products of unknown reactivity involving
the radical species formed in the electrospray ionization
process and we conclude that while 3b can perform as an
ESI tag, it is not suited to the study of systems for which
there are efficient alternative metal-based ionization path-
ways. It should be noted, however, that Pd0 is an unusually
challenging system, and that for less electron-rich metal
complexes, the idea remains valid.
The difficulty in obtaining interpretable MS data on pal-
ladium-catalyzed reactions using chargeable ESI tags led us
to the investigation of permanently charged ESI tags for the
study of these systems.[29] Permanently charged ESI tags
have the advantage of (1) being less pH-dependent than
chargeable tags, and (2) having fewer potential avenues for
reactivity that may interfere with the system of interest. In
our experience, the use of a permanently charged tag gen-
erally results in simpler spectra.[30] Ammonium or phospho-
nium groups are most commonly employed as the charged
group,[31] and if solubility of the molecule in organic sol-
vents is compromised due to the addition of this polar
group, the counterion may be exchanged to obtain more
desirable solubility properties.
[1,8-Bis(dimethylamino)naphthalene-4-yl] Bromide (2b): A solution
of bromine (2.15 mL, 41.7 mmol) in carbon tetrachloride (40 mL)
was added dropwise over 2.5 h to a stirring solution of 1 (10.0 g,
46.6 mmol) in carbon tetrachloride (60 mL) under an inert atmo-
sphere, resulting in a dark red solution. To this, aqueous solutions
of sodium thiosulfate (1 m, 20 mL) and sodium hydroxide (20% w/
v, 20 mL) were added dropwise with stirring. The resultant yellow-
brown mixture was filtered and the beige byproduct was washed
(CH2Cl2). The organic layer was collected and the volume reduced
to give 2b as a red-brown oil (7.07 g, 52% crude yield). The product
1
was used directly without further purification to synthesize 3b. H
NMR (300 MHz, CDCl3): δH = 7.80–6.71 (m, 5 H, aromatic pro-
tons), 2.81–2.77 [m, 12 H, 2(NMe2)] ppm.
[1,8-Bis(dimethylamino)naphthalene-4-yl]diphenylphosphane (3b): A
solution of 2b (1.5 g, 5.1 mmol) in THF (15 mL) was cooled to
–78 °C and nBuLi (3.3 mL, 1.6 m in hexanes, 5.1 mmol) was added
dropwise with stirring, after which the solution was stirred for a
further 30 min resulting in an amber-colored solution. Chloro-di-
phenylphosphane (1 mL, 5.4 mmol) was added dropwise with vig-
orous stirring, and stirring was continued (–78 °C, 2 h) giving a
red-orange solution. The mixture was warmed to room temperature
and the yellow-brown precipitate isolated by filtration. Orange
crystals of 3b were collected from the filtrate by solvent evaporation
(37% yield). 1H NMR (300 MHz, CDCl3): δH = 7.98 (m, 1 H,
naphthalene proton), 7.27 (m, 11 H, phenyl protons and one naph-
thalene proton), 6.90 (d, 1 H, naphthalene proton), 6.78 (m, 2 H,
naphthalene proton), 2.804 [s, 6 H, N(CH3)], 2.799 [s, 6 H,
N(CH3)] ppm. 31P NMR (360 MHz, CDCl3): δP = –13.41 [s,
PPh2(C10H17N2)] ppm. M.p. 168–169 °C.
Conclusions
Proton sponge derivatized phosphane ligands possess –
theoretically at least – many of the properties desirable for
Reaction of [Ru(η6-p-cymene)Cl2]2 with 3a in MeOH: Adapted from
a chargeable ESI-MS tag. Alone, they are selective for ion- ref.[32] A methanolic solution of 3a (2 equiv., 0.131 g, 0.33 mmol)
738
www.eurjic.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2012, 733–740

Proton Sponge Phosphanes
and [Ru(p-cymene)Cl2]2 (0.101 g, 0.165 mmol) was refluxed for 3 h crystals of trans-[PdCl2(3a)2] and a small quantity of yellow crystals
with monitoring (ESI-MS), after which time the reaction was
judged to be complete (absence of free 3a). The solution was left
to stand for 2 d, reduced in volume and layered with pentane.
Crystallizations were set up, eventually the product decomposed to
black oil. ESI-MS (after 2 h reflux, MeOH, selected peaks): m/z
(%) = 634.94 (40) [Ru(η6-p-cymene)(3a) + H]+, isotope model
(635.22), 874.53 (3) [Ru2(η6-p-cymene)(MeOH)Cl3(3a)]+, isotope
model (875.05). MS/MS (635) [Ru(η6-p-cymene)(3a) + H]+): m/z
(%) = 496.41, 457.35, 379.10, 298.81 (major Ru species) and 197.70
of [Pd2Cl4(3a)2] slowly formed at the interface. The orange crystals
proved to be highly insoluble in most solvents and only sparingly
soluble in CH2Cl2. Yellow crystals of [Pd2Cl4(3a)2] were completely
soluble in CH2Cl2. M.p. ([PdCl2(3a)2]) 178–186 °C (decomp.),
m.p. ([Pd2Cl4(3a)2]): 198 °C (decomp.). PdCl2(3a)2·CH2Cl2:
C53H56Cl4N4P2Pd: calcd. C 60.22, H 5.34, N 5.30, P 5.87; found C
59.95, H 5.39, N 5.14, P 5.64. 31P NMR (CDCl3, 202 MHz): δP
=
1
42.80 {s, [Pd2Cl4(3a)2]}, trace 32.82 (s, trans-[PdCl2(3a)2]) ppm. H
NMR (CDCl3, 500 MHz): δH = 7.81 (ddd, JHP = 13.1, JHH = 8.3,
1.0 Hz, 4 H, 12-H), 7.75 (dd, J = 7.7, 1.5 Hz, 1 H, 5-H or 7-H)
7.71 (dd, J = 8.5, 1.6 Hz, 1 H, 4-H), 7.67 (dd, J = 7.5, 1.5 Hz, 1
H, 5-H or 7-H), 7.62 (dd, J = 7.7 Hz, 1 H, 6-H), 7.57 (t, 2 H, 14-
([M
–
CH4]+, model 198.11). MS/MS (874), [Ru2Cl3(η6-p-
cymene)(3a)MeOH]+: m/z (%) = 872.56, general Ru “grass” be-
tween 800 and 400, 213.8 ([M – H]+, model 213.14).
H), 7.47 (td, JHH = 8.8, JHP = 3 Hz, 4 H, 13-H), 7.07 (dd, JHP
=
Reaction of [Ru(p-cymene)Cl2]2 with 3a in MeCN/toluene: The com-
pounds 3a (13 mg, 0.033 mmol) and [Ru(p-cymene)Cl2]2 (10 mg,
0.017 mmol) were placed in a flask, which was evacuated and
placed under N2. CD3CN (1 mL) was added, the solution mixed
9.4, JHH = 8.5 Hz, 1 H, 3-H), 4.19 (s, 6 H, NMe2), 2.62 (s, 6 H,
NMe2) ppm. Major impurity: 6.28 (m, cod, 2.5 H), 5.56 (m, cod, 2
H), 2.35 (m, cod, 4 H) ppm. ESI-MS (CH2Cl2/formic acid) selected
peaks: m/z (%) = 974.89 (0.2%, [PdCl2(3a)2 + H]+, model 975.23),
581.88 (0.35%, [MeCN·PdCl(3a)]+, model 582.08), 485.95 (0.08%,
[PdCl2(3a)2 – 2H]2+, model 486.94). MS/MS (974): m/z (%) =
974.91 ([PdCl2(3a)2 + H]+), 399.09 [3a +H]+. ESI-MS (yellow crys-
tals) (MeOH/DCM) selected peaks: m/z (%) = 1114.84 (100%,
[Pd2Cl3(3a)2]+, model 1115.10), 1078.89 (2%, [Pd2Cl2(3a) – H]+,
model 1079.12), 538.94 (100%, [PdCl·3a]+, model 539.07).
and filtered into an NMR tube. The reaction was monitored by 31
P
1
1
and H NMR at regular intervals. 31P NMR yield 80%. H NMR
(360 MHz, CD3CN, [Ru(η6-p-cymene)Cl2]2): δH = 5.54 (d, J =
6.1 Hz, 4 H, AriPr), 5.29 (d, J = 6.2 Hz, 4 H, ArMe), 2.90 (septet,
J = 7.0 Hz, 2 H, HiPr), 2.22 (s, 6 H, Me), 1.29 [d, J = 7.0 Hz, 12
H, 2(iPr)Me] ppm. 13C NMR {90 MHz, CD3CN, [Ru(η6-p-
cymene)Cl2]2}: δC = 103.52 (quat, AriPr), 99.99 (quat, ArMe), 84.63
(ArMe), 82.21 (AriPr), 31.79 (secondary iPr), 22.38 (iPrMe), 18.96
(Me) ppm. Main product, [Ru(p-cymene)(Cl)(3a)][Cl]. 31P NMR
(red crystals, 146 MHz, CD3CN): δ = 52.69 (s) ppm. 1H NMR
(360 MHz, CD3CN): δH = 7.85 [ddd, J = 11.3, 1.1, 8.3 Hz, 2 H,
H12, NOE to p-cymene (5.07)], 7.71–7.49 (m, 12 H, 1 5-H, 1 4-H,
2 12Ј-H, 2 13-H & 2 13Ј-H, 1 14-H & 1 14Ј-H, 6-H, 7-H) ppm.
Within this: 7.54–7.58 (4-H), 7.28–7.16 (m, impurity, 2 H), 7.12
(dd, 1 H, J = 8.4, 9.6 Hz, coupling to 31P, H3), 6.28 (dd, J = 7,
1 Hz, 1 H, p-cymene ring, NOE to iPr group at 1.22 and 6.07),
6.25 (dd, J = 5.2, 1 Hz, 1 H, p-cymene ring, NOE to 5.07 and Me
at 1.53), 6.07 (dd, J = 6.3, 1 Hz, 1 H, p-cymene ring, NOE to Me
group at 4.02 and Me group at 1.53), 5.07 (dd, J = 6.0, 1.5 Hz, 1
H, p-cymene ring, NOE to iPr CH group and 6.25, NOE to H12 at
7.85), 4.18 [d, J = 0.5 Hz, 3 H, NMe group, on C1back, NOE to
6.07 (weak) and 6.25 p-cymene proton], 4.02 (s, 3 H, Me, NMe
group, on C1front, NOE to 6.07 p-cymene, NOE to p-cymene Me at
1.53), 2.90 (s, 3 H, NMe group on C8back, NOE to NMe at 4.18
and weak NOE to NMe at 4.02), 2.78 (septet, J = 6.9 Hz, 1 H, iPr
CH, NOE to both 1.22 and 1.18, and 5.07), 2.27 (s, NMe group
on C8front, 3 H, NOE to Ar-CH3 at 1.53, strong NOE to NMe on
C8back at 2.90, very weak NOE to 7-H), 1.53 (s, Ar-CH3 group,
NOE to 6.25 and 6.07), 1.22 (d, J = 6.7 Hz, 3 H, iPr CH3 group,
NOE to 2.78), 1.18 (d, J = 6.9 Hz, 3 H, iPr CH3 group, NOE to
2.78). ESI-MS: m/z (%) = 669.16 (100%) ([Ru(η6-p-cymene)Cl-
(3a)]+ model 669.18). MS/MS (669): m/z (%) = 669.97 ([Ru(η6-p-
cymene)Cl(3a)]+), 536.52 ([RuCl(3a)]+ model 535.07), 482.36,
197.70 ([1a-CH3]+, model C13H13N2 197.11).
Synthesis Using 3a·HBr: ESI-MS (orange product): m/z (%) =
376.10 (1%, [“Pd”]2+), 397.19 [5%, (3a – H)]+, 399.20 [1%, (3a
+ H)]+, 486.11 (3%, [PdCl2(3a – H)2]2+, model C52H52Cl2N4P2Pd,
486.61), 575.03 (4%, [PdCl2(3a – H)]+, model C26H26Cl2N2PPd,
575.03), 607.05 (2%), [PdCl2(3a) + 31]+, 774.12 (0.5%, [Pd2Cl4(3a –
H)2(3a)]2+
, model C78H79Cl4N6P3Pd2, 774.62), 791.10 (0.2%
[“Pd”]2+), 827.05 (0.2% [“Pd2”]2+), 844.05 (0.3%, [“Pd2”]2+),
863.03 (0.3%, [“Pd2”]2+), 880.00 (3%, [Pd3Cl7(3a – H)3]2+, model
880.52), 933.45 (0.1%, [“Pd2”]2+).
Synthesis of PtCl2(3a)2: A solution of 3a (43 mg, 0.11 mmol) in
diethyl ether (1.1 mL) was added to
a stirring solution of
PtCl2(cod) (20 mg, 0.054 mmol) in CH2Cl2 (0.7 mL). The pale yel-
low solution intensified in color on stirring for 6 h after which time
stirring of the solution was stopped and ether was carefully layered
onto the solution which was left overnight. Microcrystalline yellow
PtCl2(3a)2 was recovered in two successive crops by filtration and
washing with ether: yield 23 mg, 41%, m.p. 236–240 °C (decomp).
31P NMR (400 MHz, CDCl3): δP = 15.6 (1JPPt = 3900 Hz) ppm. 1H
NMR (400 MHz, CDCl3): δH = 7.79 (m, 10 H, 8 12-H plus 2 4-
H), 7.71 (d, J = 2, 8 Hz, 2 H, 5-H or 7-H), 7.69 (d, J = 8, 2 Hz, 2
H, 5-H or 7-H), 7.62 (t, J = 7.6 Hz, 2 H, 2 6-H), 7.54 (m, 4 H, 4
14-H), 7.45 (t with 31P coupling, 8 13-H), 7.20 (dd, J = 8, 9 Hz, 2
H, 3-H), 4.40 [s, 6 H, N(CH3)2, 1-H or 8-H], 2.62 [s, 6 H,
N(CH3)2, 1-H or 8-H] ppm. ESI-MS (of crude solution, CH2Cl2):
m/z (%) = 398.89 (100%, [3a + H]+), 1062.58 (15%, [PtCl2(3a)2 +
H]+) and trace amounts (Ͻ1%) of: [PtCl(3a)2]+ (1026.65, 0.5%),
dimer at 1726.23 (model Pt2Cl4(3a)3H = 1727.38). MS/MS of
[PtCl2(3a)2 + H]+ gave exclusively free [3a + H]+ at 399.03. MS/
MS of the dimer at 1726 gave [PtCl2(3a)2 + H]+ and [3a + H]+
only.
Reaction Between [Ru(p-cymene)Cl2]2 and Protonated Ligand
3a·HBF4: The reaction was repeated by stirring at room tempera-
ture in MeCN with the protonated ligand 3a·HBF4 overnight. The
only new 31P signal detected was due to the intermediate (pos-
tulated to be the dimer) not the final product monomer. 31P NMR
(121 MHz, CD3CN): δP = –18.32 (s, 3a·HBF4, integral 35), 31.97
(s, integral 1) ppm.
CCDC-837948 (for 3a) and CCDC-837947 (for 3b) contain the
supplementary crystallographic data for this paper. These data can
be obtained free of charge from The Cambridge Crystallographic
Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
Crystalline trans-PdCl2(3a)2 (Orange) with Crystalline Pd2Cl4(3a)2
(Yellow): A yellow solution of 3a (100 mg, 0.25 mmol) in dry ether
(0.5 mL) was layered onto a solution of [Pd(COD)Cl2] (20 mg,
0.07 mmol) in dry CH2Cl2 (0.5 mL) in a narrow glass tube. Orange
Supporting Information (see footnote on the first page of this arti-
1
cle): Additional ESI-MS(/MS) and H NMR spectra.
Eur. J. Inorg. Chem. 2012, 733–740
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
739

N. J. Farrer, K. L. Vikse, R. McDonald, J. S. McIndoe
FULL PAPER
B. K. Nicholson, J. Chem. Soc., Dalton Trans. 1999, 3507–
3513.
Acknowledgments
[12] a) C. Hinderling, C. Adlhart, P. Chen, Angew. Chem. 1998, 110,
2831; Angew. Chem. Int. Ed. 1998, 37, 2685–2689; b) C.
Adlhart, C. Hinderling, H. Baumann, P. Chen, J. Am. Chem.
Soc. 2000, 122, 8204–8214; c) P. Chen, Angew. Chem. 2003,
115, 2938; Angew. Chem. Int. Ed. 2003, 42, 2832–2847.
[13] S. Zhou, B. S. Prebyl, K. D. Cook, Anal. Chem. 2002, 74, 4885–
4888.
N. J. F. thanks Professor Brian F. G. Johnson for support and the
Engineering and Physical Sciences Research Council (EPSRC) for
a studentship and for current funding (EP/G006792/1), J. S. M.
thanks the Natural Sciences and Engineering Research Council
(NSERC) of Canada for a Discovery Grant and a Discovery Accel-
erator Supplement, the Canada Foundation for Innovation (CFI)
and the British Columbia Knowledge Development Fund
(BCKDF), and the University of Victoria for instrumentation and
operational funding. K. V. thanks the University of Victoria for a
Pacific Century Fellowship.
[14] N. J. Farrer, R. McDonald, J. S. McIndoe, Dalton Trans. 2006,
4570–4579.
[15] R. W. Alder, Chem. Rev. 1989, 89, 1215–1223.
[16] a) K. Nagasawa, Related Organocatalysts (1): A Proton
Sponge, in: Superbases for Organic Synthesis: Guanidines, Ami-
dines, Phosphazenes and Related Organocatalysts (Ed.: T. Ishi-
kawa), Wiley, Chichester, U.K., 2007, pp. 931–1139.
[17] a) R. Shroff, A. Svatos, Anal. Chem. 2009, 81, 7954–7959; b)
R. Shroff, A. Svatos, Rapid Commun. Mass Spectrom. 2009,
23, 2380–2382.
[1] K. W. M. Siu, R. Guevremont, J. C. Y. Le Blanc, G. J. Gardner,
S. S. Berman, J. Chromatogr. A 1991, 554, 27–38.
[2] A. J. Canty, P. R. Traill, R. Colton, I. M. Thomas, Inorg. Chim.
Acta 1993, 210, 91–97.
[18] a) H. A. Staab, T. Saupe, Angew. Chem. 1988, 100, 895–909; b)
S. T. Howard, J. Am. Chem. Soc. 2000, 122, 8238–8244.
[19] A. F. Pozharskii, A. V. Degtyarev, O. V. Ryabtsova, V. A. Ozer-
yanskii, M. E. Kletskii, Z. A. Starikova, L. Sobczyk, A. Fi-
larowski, J. Org. Chem. 2007, 72, 3006–3019.
[20] A. L. Llamas-Saiz, C. Foces-Foces, J. Elguero, J. Mol. Struct.
1994, 328, 297–323.
[3] a) M. Bonchio, G. Licini, G. Modena, O. Bortolini, S. Moro,
W. A. Nugent, J. Am. Chem. Soc. 1999, 121, 6258–6268; b) Z.
Li, Z. H. Tang, X. X. Hu, C. G. Xia, Chem. Eur. J. 2005, 11,
1210–1216; c) B. C. Gilbert, J. R. L. Smith, A. Mairata i Pay-
eras, J. Oakes, R. Pons i Prats, J. Mol. Catal. A 2004, 219, 265–
272; d) H. Chen, R. Tagore, G. Olack, J. S. Vrettos, T.-C. Weng,
J. Penner-Hahn, R. H. Crabtree, G. W. Brudvig, Inorg. Chem.
2007, 46, 34–43.
[21] R. W. Alder, P. S. Bowman, W. R. S. Steele, D. R. Winterman,
Chem. Commun. (London) 1968, 723–724.
[4] a) P. J. Dyson, K. Russell, T. Welton, Inorg. Chem. Commun.
2001, 4, 571–573; b) P. Pelagatti, M. Carcelli, F. Calbiani, C. [22] H. M. Gilow, D. E. Burton, J. Org. Chem. 1981, 46, 2221–2225.
Cassi, L. Elviri, C. Pelizzi, U. Rizzotti, D. Rogolino, Organome-
tallics 2005, 24, 5836–5844; c) C. Daguenet, R. Scopelliti, P. J.
Dyson, Organometallics 2004, 23, 4849–4857.
[23] a) F. Hibbert, J. Emsley, Adv. Phys. Org. Chem. 1990, 26, 255–
279; b) J. P. H. Charmant, G. C. Lloyd-Jones, T. M. Peakman,
R. L. Woodward, Eur. J. Org. Chem. 1999, 2501–2510; c) J.
Berthelot, C. Guette, M. Essayegh, P. L. Desbene, J. J. Bassel-
ier, Synth. Commun. 1986, 16, 1641–1645; d) N. V. Vistorobskii,
A. F. Pozharskii, Zh. Org. Khim. 1989, 25, 2154–2161.
[24] C. A. Tolman, Chem. Rev. 1977, 77, 313–348; O. V. Ryabtsova,
A. F. Pozharskii, V. A. Ozeryanskii, N. V. Vistorobskii, Russ.
Chem. Bull. 2001, 50, 854–859.
[5] C. Vicent, M. Viciano, E. Mas-Marza, M. Sanau, E. Peris, Or-
ganometallics 2006, 25, 3713–3720.
[6] a) S. R. Wilson, Y. Wu, Organometallics 1993, 12, 1478–1480;
b) C. Raminelli, M. H. G. Prechtl, L. S. Santos, M. N. Eberlin,
J. V. Comasseto, Organometallics 2004, 23, 3990–3996; c) C.
Chevrin, J. Le Bras, F. Henin, J. Muzart, A. Pla-Quintana, A.
Roglans, R. Pleixats, Organometallics 2004, 23, 4796–4799.
[7] a) J. Masllorens, I. Gonzalez, A. Roglans, Eur. J. Org. Chem.
2007, 158–166; b) A. A. Sabino, A. H. L. Machado, C. R. D.
Correia, M. N. Eberlin, Angew. Chem. 2004, 116, 2568; Angew.
Chem. Int. Ed. 2004, 43, 2514–2518; c) P. A. Enquist, P. Nils-
son, P. Sjoberg, M. Larhed, J. Org. Chem. 2006, 71, 8779–8786;
d) A. Svennebring, P. J. R. Sjoberg, M. Larhed, P. Nilsson, Tet-
rahedron 2008, 64, 1808–1812; e) C. Chevrin, J. Le Bras, A.
Roglans, D. Harakat, J. Muzart, New J. Chem. 2007, 31, 121–
126; f) C. Markert, M. Neuburger, K. Kulicke, M. Meuwly, A.
Pfaltz, Angew. Chem. 2007, 119, 5996; Angew. Chem. Int. Ed.
2007, 46, 5892–5895.
[25] T. E. Müller, D. M. P. Mingos, Trans. Met. Chem. 1995, 20,
533–539.
[26] W. Henderson, C. Evans, Inorg. Chim. Acta 1999, 294, 183–
192.
[27] A. D. Phillips, L. Gonsalvi, A. Romerosa, F. Vizza, M. Peruz-
zini, Coord. Chem. Rev. 2004, 248, 955–993.
[28] L. S. Santos, G. B. Rosso, R. A. Pilli, M. N. Eberlin, J. Org.
Chem. 2007, 72, 5809–5812.
[29] a) K. L. Vikse, M. A. Henderson, A. G. Oliver, J. S. McIndoe,
Chem. Commun. 2010, 46, 7412–7414; b) K. L. Vikse, Z. Ah-
madi, C. C. Manning, D. A. Harrington, J. S. McIndoe, Angew.
Chem. Int. Ed. 2011, 50, 8304–8306.
[8] L. S. Santos (Ed.), Reactive Intermediates: MS Investigations in [30] a) D. M. Chisholm, A. G. Oliver, J. S. McIndoe, Dalton Trans.
Solution, Wiley-VCH, Weinheim, Germany, 2010.
[9] A. O. Aliprantis, J. W. Canary, J. Am. Chem. Soc. 1994, 116,
6985–6986.
[10] a) R. Colton, J. C. Traeger, Inorg. Chim. Acta 1992, 201, 153–
155; b) I. Ahmed, A. M. Bond, R. Colton, M. Jurcevic, J. C.
Traeger, J. N. Walter, J. Organomet. Chem. 1993, 447, 59–65.
[11] a) D. J. F. Bryce, P. J. Dyson, B. K. Nicholson, D. G. Parker,
Polyhedron 1998, 17, 2899–2905; b) C. Decker, W. Henderson,
2010, 39, 364–373 ; b) S. L. Jackson, D. M. Chisholm, J. S.
McIndoe, L. Rosenberg, Eur. J. Inorg. Chem. 2010, 327–330.
[31] D. M. Chisholm, J. S. McIndoe, Dalton Trans. 2008, 3933–
3945.
[32] C. S. Allardyce, P. J. Dyson, D. J. Ellis, S. L. Heath, Chem.
Commun. 2001, 1396–1397.
Received: August 2, 2011
Published Online: December 21, 2011
740
www.eurjic.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2012, 733–740