A retro Diels-Alder route to diphosphorus chemistry: Molecular precursor synthesis, kinetics of P2transfer to 1,3-dienes, and detection of P2by molecular beam mass spectrometry

Article abstract of DOI:10.1021/ja507922x

The transannular diphosphorus bisanthracene adduct P₂A₂(A = anthracene or C₁₄H₁₀) was synthesized from the 7-phosphadibenzonorbornadiene Me₂NPA through a synthetic sequence involving chlorophosphine ClPA (28-35%) and the tetracyclic salt [P₂A₂Cl][AlCl₄] (65%) as isolated intermediates. P₂A₂was found to transfer P₂efficiently to 1,3-cyclohexadiene (CHD), 1,3-butadiene (BD), and (C₂H₄)Pt(PPh₃)₂to form P₂(CHD)₂(>90%), P₂(BD)₂(69%), and (P₂)[Pt(PPh₃)₂]₂(47%), respectively, and was characterized by X-ray diffraction as the complex [CpMo(CO)₃(P₂A₂)][BF₄]. Experimental and computational thermodynamic activation parameters for the thermolysis of P₂A₂in a solution containing different amounts of CHD (0, 4.75, and 182 equiv) have been obtained and suggest that P₂A₂thermally transfers P₂to CHD through two competitive routes: (i) an associative pathway in which reactive intermediate [P₂A] adds the first molecule of CHD before departure of the second anthracene, and (ii) a dissociative pathway in which [P₂A] fragments to P₂and A prior to addition of CHD. Additionally, a molecular beam mass spectrometry study on the thermolysis of solid P₂A₂reveals the direct detection of molecular fragments of only P₂and anthracene, thus establishing a link between solution-phase P₂-transfer chemistry and production of gas-phase P₂by mild thermal activation of a molecular precursor.

Full text of DOI:10.1021/ja507922x

Communication
pubs.acs.org/JACS
A Retro Diels−Alder Route to Diphosphorus Chemistry: Molecular
Precursor Synthesis, Kinetics of P Transfer to 1,3-Dienes, and
2
Detection of P by Molecular Beam Mass Spectrometry
2
†
†
‡
†
†
Alexandra Velian, Matthew Nava, Manuel Temprado, Yan Zhou, Robert W. Field,
,†
and Christopher C. Cummins*
†
Department of Chemistry, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, United States
‡
Department of Analytical Chemistry, Physical Chemistry and Chemical Engineering, Universidad de Alcala,
́
Madrid 28871, Spain
*
S Supporting Information
Scheme 1. Synthesis of 1 (P A , A = C H or Anthracene)
2
2
14 10
ABSTRACT: The transannular diphosphorus bisanthra-
cene adduct P2A2 (A = anthracene or C H ) was
Starting from Me NPCl and MgA·3THF
2
2
14
10
synthesized from the 7-phosphadibenzonorbornadiene
Me NPA through a synthetic sequence involving chloro-
2
phosphine ClPA (28−35%) and the tetracyclic salt
[
P A Cl][AlCl ] (65%) as isolated intermediates. P A
2 2 4 2 2
was found to transfer P efficiently to 1,3-cyclohexadiene
2
(
CHD), 1,3-butadiene (BD), and (C H )Pt(PPh ) to
2 4 3 2
form P (CHD) (>90%), P (BD) (69%), and (P )[Pt-
2
2
2
2
2
(
PPh ) ] (47%), respectively, and was characterized by X-
3 2 2
ray diffraction as the complex [CpMo(CO) (P A )][BF ].
3
2
2
4
Experimental and computational thermodynamic activa-
tion parameters for the thermolysis of P A in a solution
2
2
containing different amounts of CHD (0, 4.75, and 182
equiv) have been obtained and suggest that P A thermally
2
2
transfers P to CHD through two competitive routes: (i)
2
an associative pathway in which reactive intermediate
[
P2A] adds the first molecule of CHD before departure of
the second anthracene, and (ii) a dissociative pathway in
which [P A] fragments to P and A prior to addition of
2
2
CHD. Additionally, a molecular beam mass spectrometry
study on the thermolysis of solid P A reveals the direct
2
2
detection of molecular fragments of only P2 and
anthracene, thus establishing a link between solution-
so obtained, we sought to access P by a photochemical route,
2
reasoning that if indeed P is a common reactive intermediate,
2
phase P -transfer chemistry and production of gas-phase P2
2
then the same trapping products should result irrespective of the
by mild thermal activation of a molecular precursor.
P source. Our work on direct incorporation of P units into
2
2
organic molecules via UV irradiation of P /1,3-diene mixtures
4
8
ensued, and theoretical studies suggest accordingly that a direct
dissociation pathway exists for P photo-cracking to 2P .
Our recent report of easy access to the 7-phosphadibenzo-
norbornadiene skeleton now opens up a new possibility for
9
n important theme in modern main-group chemistry is
4
2
10
A
concerned with multiple bonds between the heavier
1
,2
members of the p-block elements. Noteworthy examples of
unusual homo-diatomic triple bonds that could be stabilized
using sterically demanding substituents or donor ligands include
thermal molecular precursor access to P chemistry, one that is in
2
line with traditional retro Diels−Alder routes for reactive
3
11
the diboryne reported by Braunschweig et al. and the disilyne
intermediate generation. Magnesium anthracene (MgA·
4
reported by Sekiguchi et al. In this vein, we have been inspired to
3THF, A = C H or anthracene) reaction with Cl PNMe
14
10
2
2
seek out access to the solution-phase chemistry of the triply
proceeds with formation of a 7-phosphadibenzonorbornadiene
bearing a dimethylamino substituent at phosphorus (Scheme 1).
Dimethylamino group replacement is effected by HCl treatment,
providing the corresponding chloro-substituted 7-phospha-
bonded P molecule by generating it as a reactive intermediate.
2
Along these lines, we developed initially a niobium-complexed
diphosphaazide ligand (PPNMes*, diphosphorus analogue
12
dibenzonorbornadiene (ClPA, see Scheme 1 and Figure 1).
of an organic azide) that was a molecular source of P , effecting
2
transfer of the diphosphorus molecule to a variety of organic and
5
,6
inorganic acceptors upon mild thermal activation. Intrigued by
Received: August 2, 2014
the novel nature of several of the P -containing reaction products
Published: September 8, 2014
2
©
2014 American Chemical Society
13586
dx.doi.org/10.1021/ja507922x | J. Am. Chem. Soc. 2014, 136, 13586−13589

Journal of the American Chemical Society
Communication
Scheme 2. Transfer of P from 1 (P A ) to 1,3-Cyclohexadiene
2
2 2
(
CHD) and 1,3-Butadiene (BD) and to (C H )Pt(PPh ) ,
2 4 3 2
Complexation of 1 with [Mp][BF ] (Mp = CpMo(CO) ), and
4
3
Thermolysis of 1 in Solution and in the Solid State
Figure 1. Solid-state molecular structures of ClPA and [P A Cl][AlCl ]
2
2
4
7
(
[
50% ellipsoids, 100 K) rendered using PLATON, with H-atoms, the
AlCl ] counterion, and the disordered solvent (see SI S.5) omitted for
−
4
clarity.
The coupling together of two 7-phosphadibenzonorborna-
diene groups occurred upon treatment of two ClPA with one
1
3
14
AlCl3, and subsequent rearrangement via an observable
intermediate (see Supporting Information (SI) S.1.10 and
S.1.11) resulted in a tetracyclic skeleton harboring a P unit at
2
the junction of two fused dibenzodiphosphabicycloocta[2.2.2]-
diene bicycles. One P−Cl bond still remained in this
+
phosphonium cation, [P A Cl] , isolated and structurally
2
2
characterized as its tetrachloroaluminate salt (Scheme 1 and
Figure 1). All that remained was to reductively dechlorinate
+
15
cation [P A Cl] by treatment with LiAlH , in order to afford
2
2
4
the desired neutral P precursor, P A (1, title molecule boxed in
2
2 2
1
6
Scheme 1), effectively a metastable sandwich of phosphorus (as
P2) between two anthracene molecules.
As expected under the assumption that compound 1 can serve
as a thermal P source, heating 1 in the presence of excess 1,3-
2
cyclohexadiene (CHD) leads to high-yield formation of
Figure 2. Solid-state molecular structure of [1-Mp][BF ] (Mp =
tetracyclic P (CHD) (Scheme 2), a compound first encoun-
4
2
2
7
5
CpMo(CO) ) (50% ellipsoids, 100 K) and rendered using PLATON,
tered using niobium diphosphaazide as the thermal P source.
3
2
−
with H-atoms and the [BF ] counterion omitted for clarity.
4
Heating compound 1 in the presence of excess unsubstituted 1,3-
butadiene (BD) delivered the corresponding bicyclic derivative,
P (BD) , in good yield for the first time (Scheme 2); it was
2
2
previously isolated in trace quantities after being generated via
diphosphene [P A], can either (eq 5) lose a second anthracene
2
the photochemical P -cracking route and structurally charac-
4
molecule by retro Diels−Alder to generate P or (eq 2) undergo
2
8
terized.
[
4+2] cycloaddition to CHD to generate the observable
Heating compound 1 in a benzene solution in the absence of
an added trapping agent led to quantitative generation of
anthracene, formation of an orange phosphorus precipitate and,
intermediate P A(CHD). For computational consideration of
2
further conceivable reactions of the reactive intermediate [P A],
2
such as dimerization or reaction with P en route to P , the reader
2
4
interestingly, an appreciable amount of dissolved P (see S.3.1).
4
is directed to SI Figure S.28.
It should be noted also that compound 1, which is colorless when
pure, changes color to orange due to production of a form of
polymeric red phosphorus upon exposure to light, with
Kinetic study of the P -transfer reaction, monitored using
2
phosphorus and proton NMR spectroscopy, has provided first-
order rate constants for the decay of 1 as a function of
temperature, listed in SI Table S.1. Importantly, the decay rate for
17
formation of anthracene.
While we have not yet obtained the crystal structure of 1 by
1
was found to be zero order in the concentration of CHD. To
itself, we found that 1 forms a crystalline Mp+ (Mp =
better understand the thermal transfer of P to CHD, we
2
CpMo(CO) ) derivative in which the tetracyclic skeleton is
3
undertook kinetic modeling of the process accounting
simultaneously for decay of 1, accumulation and decay of
intact (Scheme 2 and Figure 2). Interestingly, the thermolysis of
a pyridine-d solution of [1-Mp][BF ] revealed the formation of
5
4
observable intermediate P A(CHD), and formation of product
2
18
P and the previously reported CpMo(CO) P complex, along
4
2
3
P (CHD) . In the regime of high diene concentration, the
2
2
with anthracene. Treating 1 with the platinum ethylene complex
simultaneous time evolution of all three observable species is
accounted for nicely according to eqs 1−4:
6
(
C H )Pt(PPh ) , a competent acceptor of diphosphorus, led
2 4 3 2
to the formation of the previously reported platinum
19
k1
diphosphorus complex (P )[Pt(PPh ) ] at room temperature,
1 → [P A] + A
2
3 2 2
(1)
2)
2
with release of anthracene and ethylene.
k
2
The complete transfer of P from 1 to CHD has been the
2
[P A] + CHD → P A(CHD)
(
2
2
subject of a kinetic investigation and a computational study
k
3
(
Figures 3 and 4; eqs 1−6 below). The quantum chemical
calculations suggest that retro Diels−Alder loss of anthracene
with formation of a PP double bond) is rate-determining.
The first reactive intermediate thus formed, unobserved cyclic
P A(CHD) → [P (CHD)] + A
(3)
(4)
2
2
(
k
4
[
P (CHD)] + CHD → P (CHD)
2
2
2
1
3587
dx.doi.org/10.1021/ja507922x | J. Am. Chem. Soc. 2014, 136, 13586−13589

Journal of the American Chemical Society
Communication
Figure 3. Combined DFT-calculated potential energy diagram (M06-2X, 6-311G(3df,2p), IEF-PCM in THF) for the thermolysis of P A in the
2
2
presence of 1,3-cyclohexadiene (D).
In modeling 1−4, loss of anthracene and addition of CHD
were treated as irreversible, and unobserved diphosphene
reactive intermediates [P A] and [P (CHD)] were treated
2
2
20
with the steady-state approximation. In the regime of low diene
concentration, however, this simple linear model sequence was
not satisfactory. Since low diene concentration would make
reaction 2 less competitive with the alternative channel of second
anthracene loss (reactions 5 and 6),
k
5
[
P A] → [P ] + A
(5)
(6)
2
2
k
6
[
P ] + CHD → [P (CHD)]
2
2
we added this channel to the model, resulting in a good fit of the
simultaneous time evolution of all three observable species. The
channel of reaction 5, “the P pathway”, is suggested by the
2
computational study to have a similar activation barrier in
comparison with the competing pathway represented by reaction
2
. The experimentally determined branching ratio (c, the
microscopic rate of reaction 5 divided by that of reaction 2) is
plotted in Figure 4 as a function of time and temperature in this
regime of low diene concentration, where it applies. Note the
Figure 4. Plot displaying the branching ratio c = k /(k [D]) at
5
2
temperatures ranging from 309 to 346 K, as a function of the logarithm
of time (in seconds) for the thermolysis of P A in THF in the presence
of a small excess of 1,3-cyclohexadiene (D), with the initial molar
2
2
increasing dominance of the unimolecular P pathway (reaction
2
5
) over bimolecular reaction 2 at elevated temperature,
concentration [D] = 89.16 mM and ratio [D] /[P A ] = 4.75.
0
0
2 2 0
consistent with expectations based upon entropy considerations.
There is precedent for the thermal delivery of small-molecule
reactive intermediates into the gas phase by retro Diels−Alder
processes. Corey and Mock reported that, upon heating of a solid
sample of the diimide (HNNH) anthracene Diels−Alder
adduct, it underwent decomposition “with evolution of a gas but
tubing connection between TGA and MS, in transit through
which P would need to survive many collisions.
2
Gradual heating of solid 1 on a temperature-controlled stage
located within the source vacuum chamber of a pulsed-jet
molecular beam instrument equipped with mass-spectrometric
21
without melting to give a solid residue of anthracene”. In a
seminal report, the related Diels−Alder adduct of nitroxyl
23
detection (MBMS) led to the successful detection of gas-phase
24
(
HNO) with 9,10-dimethylanthracene was used to deliver
P (62 m/z), P (31 m/z), and a fragment of anthracene (89 m/
2 1
HNO into the gas phase, where it was definitively characterized
z), but not of P , P , or [P A] fragments (Figure 5). By design,
3 4 2
the molecular beam of the MBMS apparatus ensures a collision-
free, post-thermolysis propagation of the molecular fragments to
22
by microwave spectroscopy. Given this precedent, we were
interested to discover whether mild thermal activation of solid
samples of compound 1 would make possible the detection of
diphosphorus in the gas phase.
the mass spectrometer; the detection of P from 1 thus implies
2
that P forms in a single, unimolecular event. Upon completion of
2
Thermogravimetric analysis (TGA) of solid 1 evinced a mass
the MBMS experiment, an orange residue was left inside of the
sample holder. Its analysis using Raman spectroscopy revealed
the formation of red phosphorus, while its combustion analysis
change consistent with P loss at 120 °C, a promising preliminary
2
indication (Figure S.18). Heating solid samples of 1 in a
sublimator resulted in pure anthracene on the cold probe and a
red coating of phosphorus around the top of the apparatus near
17,25
indicated ∼70% phosphorus content.
To check if under
identical experimental conditions of temperature (20−150 °C)
−7
the vacuum exit port (Figure S.18). Attempts to detect P in a
and pressure (∼8 × 10 Torr) red phosphorus can generate P ,
2
2
combined TGA mass spectrometry (TGA-MS) study of 1 were
inconclusive, likely due to the small-diameter, long pathlength
we proceeded to analyze a sample of commercially available red
2
6
phosphorus using our MBMS apparatus. However, no
1
3588
dx.doi.org/10.1021/ja507922x | J. Am. Chem. Soc. 2014, 136, 13586−13589

Journal of the American Chemical Society
Communication
(5) Piro, N. A.; Figueroa, J. S.; McKellar, J. T.; Cummins, C. C. Science
2
006, 313, 1276.
(6) Piro, N. A.; Cummins, C. C. Inorg. Chem. 2007, 46, 7387.
(
(
(
7) Spek, A. L. J. Appl. Crystallogr. 2003, 36, 7.
8) Tofan, D.; Cummins, C. C. Angew. Chem., Int. Ed. 2010, 49, 7516.
9) Wang, L.-P.; Tofan, D.; Chen, J.; Van Voorhis, T.; Cummins, C. C.
RSC Adv. 2013, 3, 23166.
(10) Velian, A.; Cummins, C. C. J. Am. Chem. Soc. 2012, 134, 13978.
(11) Rickborn, B. Organic Reactions; John Wiley & Sons, Inc.: New
York, 2004.
12) Mercier, F.; Deschamps, B.; Mathey, F. J. Am. Chem. Soc. 1989,
11, 9098.
13) (a) Cowley, A. H.; Kemp, R. A. Chem. Rev. 1985, 85, 367.
b) Burford, N.; Ragogna, P. J.; McDonald, R.; Ferguson, M. J. J. Am.
Chem. Soc. 2003, 125, 14404.
14) Dillon, K. B.; Mathey, F.; Nixon, J. F. Phosphorus: the carbon copy:
(
1
(
(
(
from organophosphorus to phospha-organic chemistry, 1st ed.; Wiley: West
Figure 5. Plot displaying the measured current intensity of P fragments
x
Susex, 1998.
(15) A precedented transformation,
3
2,33
(
x = 1−4), [P A], and anthracene (C H ) as a function of time and
2
14 10
the reduction of chloro-
temperature, corrected for background signal intensity, and measured
using a molecular beam mass spectrometry (MBMS) apparatus.
phosphanium salts with LiAlH is known to produce phosphines, H ,
4
2
and complex salts of formula Li AlCl H (x = 0−1, y = 0−4).
x
y
4−y
(
16) P A displays a chemical shift of −42 ppm in its 1P NMR
3
2
2
spectrum, which shift compares well to those of other polycyclic
diphosphanes. Isolated as a microcrystalline white solid sparingly
detectable amounts of P fragments were formed in the gas phase
x
8
during the experiment. The unambiguous detection of P from
2
soluble in benzene, but soluble in tetrahydrofuran, P A is a light-
2
2
thermolysis of 1, corroborated with the lack of P detection upon
2
sensitive material.
heating red phosphorus under similar conditions, provides
decisive evidence that 1 is indeed a thermal molecular precursor
of P2.
(
17) It has been reported that red phosphorus forms upon the
9,25
condensation of diphosphorus molecules. Also, detailed in SI S.4 is
our computational investigation on P formation by thermolysis of a
4
The generation of reactive intermediates by retro Diels−Alder
solution of P A in benzene, showing that it could proceed via the
2
2
chemistry has previously afforded access to the solution-phase
dimerization of [P A].
2
2
7
chemistry of interesting molecular species, including singlet O ,
̈
(18) Scherer, O.; Sitzmann, H.; Wolmershauser, G. J. Organomet.
Chem. 1984, 268, C9.
2
2
8,29
22
21
30
singlet S ,
HNO, diazene, dimethyldisilyne, and
with the synthesis of 1, we can now add P to this
2
1
0,31
(
19) Schaf
̈
er, H.; Binder, D. Z. Anorg. Allg. Chem. 1988, 560, 65.
others;
2
(20) (a) Espenson, J. Chemical Kinetics and Reaction Mechanisms;
list. The present work illustrates an exciting nexus of main-group
element multiple-bond chemistry with that of reactive
intermediates; we expect that this approach can be extended
further to reactive intermediates involving novel combinations of
main-group elements for study of reactivity patterns, evaluation
of fundamental properties, and applications in synthesis.
McGraw-Hill: New York, 1981. (b) Moore, J. W.; Pearson, R. G. Kinetics
and Mechanism; John Wiley & Sons: New York, 1981.
(
(
21) Corey, E. J.; Mock, W. L. J. Am. Chem. Soc. 1962, 84, 685.
22) Corrie, J. E. T.; Kirby, G. W.; Laird, A. E.; Mackinnon, L. W.;
Tyler, J. K. J. Chem. Soc., Chem. Commun. 1978, 275.
23) (a) Hansen, N.; Cool, T. A.; Westmoreland, P. R.; Kohse-
Honghaus, K. Prog. Energy Combust. Sci. 2009, 35, 168. (b) Kroto, H. W.;
Heath, J. R.; O’Brien, S. C.; Curl, R. F.; Smalley, R. E. Nature 1985, 318,
62.
24) (a) Gingerich, K. A. J. Phys. Chem. 1964, 68, 768. (b) Gingerich, K.
A. J. Chem. Phys. 1966, 44, 1717.
(
̈
ASSOCIATED CONTENT
Supporting Information
Synthetic, kinetic, spectroscopic, computational, and crystallo-
graphic details. This material is available free of charge via the
Internet at http://pubs.acs.org.
■
*
S
1
(
(25) Melville, H. W.; Gray, S. C. Trans. Faraday Soc. 1936, 32, 271.
(26) (a) Kerwin, L. Can. J. Phys. 1954, 32, 757. (b) Carette, J.-D.;
Kerwin, L. Can. J. Phys. 1961, 39, 1300. (c) Kane, J. S.; Reynolds, J. H. J.
Chem. Phys. 1956, 25, 342.
(27) Moureu, C.; Dufraisse, C.; Dean, P. M. C. R. Acad. Sci., Paris 1926,
182, 1440.
AUTHOR INFORMATION
Corresponding Author
ccummins@mit.edu
Notes
■
(
28) Wasserman, H. H.; Scheffer, J. R.; Cooper, J. L. J. Am. Chem. Soc.
972, 94, 4991.
29) (a) Ando, W.; Sonobe, H.; Akasaka, T. Tetrahedron Lett. 1987, 28,
653. (b) Ando, W.; Sonobe, H.; Akasaka, T. Tetrahedron 1990, 46,
5093.
1
(
6
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This material is based upon work supported by the National
Science Foundation under CHE-1362118 and the Spanish
Ministry of Economy and Competitiveness (MINECO) under
CTQ2012-36966 (M.T.). Ioana Knopf is thanked for a donation
of [Mp][BF4].
(30) Sekiguchi, A.; Zigler, S. S.; West, R.; Michl, J. J. Am. Chem. Soc.
1986, 108, 4241.
(31) Neumann, W. P.; Schriewer, M. Tetrahedron Lett. 1980, 21, 3273.
(32) Byrne, P. A.; Rajendran, K. V.; Muldoon, J.; Gilheany, D. G. Org.
Biomol. Chem. 2012, 10, 3531.
(33) (a) Finholt, A. E.; Bond, A. C.; Schlesinger, H. I. J. Am. Chem. Soc.
1947, 69, 1199. (b) Nystrom, R. F.; Brown, W. G. J. Am. Chem. Soc.
1947, 69, 2548. (c) Nystrom, R. F.; Brown, W. G. J. Am. Chem. Soc.
1947, 69, 1197. (d) Nystrom, R. F.; Brown, W. G. J. Am. Chem. Soc.
1948, 70, 3738.
REFERENCES
1) Wang, Y.; Robinson, G. H. Chem. Commun. 2009, 5201.
2) Power, P. P. Organometallics 2007, 26, 4362.
3) Braunschweig, H.; Dewhurst, R. D.; Hammond, K.; Mies, J.;
■
(
(
(
Radacki, K.; Vargas, A. Science 2012, 336, 1420.
(
4) Sekiguchi, A.; Kingo, R.; Ichinohe, M. Science 2004, 305, 1755.
1
3589
dx.doi.org/10.1021/ja507922x | J. Am. Chem. Soc. 2014, 136, 13586−13589