The substituent effect of the phosphaalkenyl group

Article abstract of DOI:10.1002/ejoc.200200618

Several novel phosphaalkenes (E)-Mes*P=CHC₆H₄X (6; X = Me₃Si, Me₃Sn, CHO, COOH, CN) were prepared and, together with other known members of this series, subjected to an analysis of linear free-energy relationships based on the values of pK_a and values of δ(¹³C_para) of substituted benzoic acids and on the values of δ(¹³C_para) of substituted bromobenzenes. These analyses indicate that the (E)-es*P=CH group is a weak electron donor with a predominantly inductive effect (δp = -0.1) on a neighboring benzene ring; resonance interactions are of minor importance. This result shows that the P=C bond is surprisingly analogous to the C=C bond, and thus confirms the well-known diagonal relationship in the Periodic System between carbon and phosphorus atoms. The relationship between X and the values of δ(³¹P) of 6 is quite pronounced. It confirms the dominance of inductive interactions on the value of δ(³¹P); only with strongly electron-donating substituents may a certain degree of resonance interaction be involved. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2003.

Full text of DOI:10.1002/ejoc.200200618

FULL PAPER
The Substituent Effect of the Phosphaalkenyl Group
Arjan Termaten,[a][‡] Marcel van der Sluis,[a][‡] and Friedrich Bickelhaupt*[a]
Keywords: Linear free energy relationships / NMR spectroscopy / Phosphaalkenes
Several novel phosphaalkenes (E)-Mes*P=CHC6H4X (6;
X = Me3Si, Me3Sn, CHO, COOH, CN) were prepared and,
together with other known members of this series, subjected
to an analysis of linear free-energy relationships based on
result shows that the P=C bond is surprisingly analogous to
the C=C bond, and thus confirms the well-known diagonal
relationship in the Periodic System between carbon and
phosphorus atoms. The relationship between X and the
the values of pKa and values of δ(13Cpara) of substituted values of δ(31P) of 6 is quite pronounced. It confirms the dom-
benzoic acids and on the values of δ(13Cpara) of substituted inance of inductive interactions on the value of δ(31P); only
bromobenzenes. These analyses indicate that the (E)- with strongly electron-donating substituents may a certain
Mes*P=CH group is a weak electron donor with a predomin- degree of resonance interaction be involved.
antly inductive effect (σp = −0.1) on a neighboring benzene
ring; resonance interactions are of minor importance. This
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2003)
Introduction
The classical double bond rule states that elements of the
third and higher periods lack the ability to form stable com-
pounds involving pπϪpπ bonds analogous to the CϭC
double bond of alkenes.[1,2] According to this rule, phos-
phaalkenes R1PϭCR2R3 (1) are expected to be chemically
quite different from alkenes, and indeed they are, for ex-
ample, as far as ease of preparation or thermal stability are
concerned; special precautions such as steric protection or
resonance stabilization of the PϭC bond are required to
ensure their viability under ordinary conditions.[3Ϫ5]
Another potential difference between the two types of
double bonds is their electronic structure, in particular the
polarization. Apart from substituent effects, the CϭC bond
is inherently nonpolar. For phosphaalkenes, the situation
may be different because the two elements differ in elec-
tronegativity χ,[6Ϫ9] with phosphorus being more electro-
positive than carbon [on the Pauling scale:[6] χ(P) ϭ 2.1,
χ(C) ϭ 2.5]. Consequently, one would predict 1 to be polar-
ized as indicated by the resonance structure 1A and not
according to 1B (Scheme 1). In line with this assumption,
phosphaalkenes react with polar reagents by attack of a nu-
cleophile at the phosphorus atom or of an electrophile at
the carbon atom;[3Ϫ5] this regiochemistry is opposite to that
of CϭO and CϭN bonds as oxygen and nitrogen atoms are
more electronegative than carbon atoms.
Scheme 1
To clarify the question of the polarity of the PϭC bond,
we had earlier investigated the 31P and 13C chemical shifts
of several triarylphosphaalkenes[10,11] 2 as well as NMR,
UV, and IR spectra and X-ray crystal structures of β-phos-
phaenones[12] 3 (Scheme 2). For 2 (Mes ϭ 2,4,6-trimeth-
ylphenyl), the values of δ(13Cpara) of the aryl carbon atoms
para to the MesPϭC group (shown in Scheme 2) suggest
that this group has a Hammett σ value of ca. 0.2. This
result was unexpected as it would mean that contrary to the
‘‘natural’’ polarization as indicated by resonance structure
1A, the group is slightly electron-withdrawing; this
phenomenon would be more in line with 1B. A calcu-
lational analysis[12] of the results obtained for 3 (Mes* ϭ
[a]
Scheikundig Laboratorium, Vrije Universiteit,
De Boelelaan 1083, 1081 HV Amsterdam, The Netherlands
Fax: (internat.) ϩ 31-20/444-7488
E-mail: bicklhpt@chem.vu.nl
[‡]
Present address: Syncom B.V., Kadijk 3, 9747 AT Groningen,
The Netherlands
Scheme 2
Eur. J. Org. Chem. 2003, 2049Ϫ2055
DOI: 10.1002/ejoc.200200618
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2049

A. Termaten, M. van der Sluis, F. Bickelhaupt
FULL PAPER
2,4,6-tri-tert-butylphenyl ϭ supermesityl), and of the spec-
tacular shielding[11] of the phosphorus atom in 4 [δ(31P) ϭ
89.9] as compared to 5 [δ(31P) ϭ 189.9], revealed that sev-
eral factors play a role and that the situation is far from
clear-cut.
For this reason, we decided to extend the range of phos-
phaalkenes to investigate additional Hammett-type linear
free-energy relations.[13Ϫ16]
Results and Discussion
Scheme 4
We concentrated our investigations on the phosphaal-
kenes 6 that are rather stable because they are protected by
the bulky supermesityl group Mes* at the phosphorus atom
(Scheme 3; they are arranged in the order of increasing σp
of the para substituent X). Correlations were derived with
Compound 6j served as starting material for the synthesis
of 6e,g,k,l as shown in Scheme 5. In THF at Ϫ65 °C, 6j was
treated with n-butyllithium, and the resulting aryllithium
intermediate was quenched with electrophilic reagents E
(E ϭ ClSiMe3, ClSnMe3, dimethyl formamide, or CO2,
respectively). Again, the yields were high (about 95%).
ϩ
n
Hammett σp or σp values[15,16] and with σp values[16Ϫ19]
for values of pKa of benzoic acids, for the values of
δ(13Cpara) of the para-carbon atoms of substituted benzoic
acids (6l and 7) and of bromobenzenes (6j and 8), and for
the values of δ(31P) of the PϭC bond of 6.
Scheme 5
Linear Free-Energy Relationships
The original Hammett Equation (1)[13] defines the influ-
ence of substituents on the acidity of benzoic acids where
K is the dissociation constant of the acid, σX is a measure
of the electron-donating or -withdrawing effect of the sub-
stituent X, and ρ is the reaction constant; by definition, ρ ϭ
1 in aqueous solution. In general, ρ is a measure of the
sensitivity of the property under investigation towards sub-
stituent effects.
Scheme 3
Synthesis of (E)-Mes*P؍CHAr (6)
σX·ρ ϭ logKX Ϫ logKH
(1)
Several of the phosphaalkenes 6 have already been pre-
For this reason, we first determined the acidity of substi-
pared by a number of different approaches.[3Ϫ5] This obser- tuted benzoic acids 7. As most of these acids were insuf-
vation holds for the unsubstituted parent compound ficiently soluble in water, the value of pKa was determined
6f[20Ϫ23] and for 6a,[21Ϫ23] 6b,[22] 6c,[21,22,24] 6d,[23] 6g,[23] in 85.6% ethanol/water by titration with NaOH (Table 1).
6h,[22] 6i,[22,24] 6j,[22] and 6o.[24] Using the method of Yoshi- It should be noted that the value of pKa is strongly depen-
fuji et al.,[20] we obtained 6j and 6n in a one-pot procedure dent on the solvent;[15] in more polar solvents, the dis-
by conversion of Mes*PH2 (9) via the silyl derivative 10 into sociation constant is increased, as is illustrated by the values
its lithio derivative 11, which was treated with p-bromo- or for the unsubstituted parent, benzoic acid (7f), which are
p-cyanobenzaldehyde to furnish 6j or 6n, respectively, in pKa(water[25]) ϭ 4.19 and pKa(85.6% ethanol/water) ϭ 6.88.
yields of 90Ϫ95% (Scheme 4). This reaction was stereospec- Similarly, the reaction constant for the dissociation is (by
ific in providing the (E) isomer exclusively.
definition[13]) ρ ϭ 1 in water, and ρ ϭ 1.57 in ethanol.[26]
2050
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org Eur. J. Org. Chem. 2003, 2049Ϫ2055

The Substituent Effect of the Phosphaalkenyl Group
FULL PAPER
dence against their validity in the present context. These
observations confirm the conclusion that the dissociation
of benzoic acids is largely determined by general inductive
effects of the substituent X and that resonance effects are
of minor importance.[15,16]
Insertion of pKa(6l) ϭ 6.88 into Equation (2) yields
σp(Mes*PϭCH) ϭ Ϫ0.12. This result was surprising as it
characterizes the phosphaalkenyl group as a weak electron
donor, which is in line with its ‘‘reactivity pattern’’ and with
the resonance structure 1A mentioned in the Introduction,
but it is opposite to the value of σ ϭ 0.2 reported[11,12] for
Table 1. Values of pKa of benzoic acids 7 in 85.6% ethanol/water
ϩ
Compound
X
σp
σp
σn
pKa
7a
7b
7c
7d
7f
7h
7i
7j
NMe2
OH
OMe
Me
H
F
Cl
Br
CF3
CN
Ϫ0.83
Ϫ0.37
Ϫ0.27
Ϫ0.17
0.00
0.06
0.23
0.23
0.54
0.66
Ϫ1.70
Ϫ0.92
Ϫ0.78
Ϫ0.31
0.00
Ϫ0.07
0.11
0.15
0.61
0.66
Ϫ0.17
Ϫ0.18
Ϫ0.11
Ϫ0.13
0.00
0.06
0.24
0.26
0.53
0.67
7.88
7.45
7.15
7.13
6.96
6.62
6.44
6.39
6.02
7m
7n
5.69 2 and to a contribution by structure 1B.
Another indicator that has been used frequently to deter-
mine substituent effects are NMR spectroscopic chemical
shifts, in particular values of δ(13Cpara) of the carbon atoms
The Hammett plot for 85.6% ethanol/water yields Equa-
para
to
the
substituent
X
(see
Scheme 2;
n
tions (2), (3), and (4) for σp, σpϩ, and σp , respectively,
Table 2).[14Ϫ16,27Ϫ30] For the benzoic acids 6l and 7, the val-
ues were obtained in [D6]DMSO because many compounds
7 are insufficiently soluble in CDCl3; spectra of the bromo-
benzenes 6j and 8 were measured in CDCl3.
where σp is the ‘‘normal’’ substituent constant for the dis-
sociation of benzoic acids; it is mainly dependent on induc-
tive effects of the substituent,[15,16] whereas σpϩ is appropri-
ate in the case of through-conjugation, when strong electron
demand is involved.[15,16] In contrast, σn has been derived
for systems that lack conjugative interactions and is sup-
posed to reflect purely inductive effects.[16,19] Only the plot
for σp is shown in Figure 1.
Like the values of pKa, the correlations of the values of
δ(13Cpara) are sensitive to solvent polarity and solvation,
factors that determine the relative contribution of inductive
effects, normally correlated with σp values, and resonance
effects, which correlate with σpϩ. Empirically, the corre-
ϩ
lation of values of δ(13Cpara) with those of σp appears to
give the best fit,[14,28Ϫ30] but there is extensive discussion
in the literature concerning the reasons for this empirical
observation.[16] From the data in Table 2, we, too, find the
ϩ
best fit for σpϩ. Equation (5) gives the correlation of σp
for 7 (Figure 2) and Equation (6) for 8 (Figure 3). The cor-
relations for σp were less satisfactory (R ϭ 0.93 and 0.92,
respectively).
7: δ(13Cpara) ϭ 129.36 ϩ 8.09σpϩ (R ϭ 0.977)
8: δ(13Cpara) ϭ 122.10 ϩ 9.10σpϩ (R ϭ 0.981)
(5)
(6)
When inserted into Equation (5), the value for 6l
[δ(13Cpara) ϭ 128.54 ppm] furnishes σpϩ(Mes*PϭCH) ϭ
Ϫ0.12, and similarly from Equation (6), 6j [δ(13Cpara) ϭ
121.58 ppm] yields σpϩ(Mes*PϭCH) ϭ Ϫ0.06. These val-
ues are in surprisingly good agreement with σp(Mes*Pϭ
CH) ϭ Ϫ0.12 from the pKa experiments. The values are
rather small, however, and if one considers the (less re-
liable!) σp values (6l: 0.09; 6j: 0.11), one comes to the con-
clusion that the electronic effect of the phosphalkenyl group
is slightly electron-donating, but close to zero. A compari-
Figure 1. Hammett plot for values of pKa of 7
pKa ϭ 6.78 Ϫ 1.49σp (R ϭ 0.982)
pKa ϭ 6.57 Ϫ 0.88σpϩ (R ϭ 0.919)
pKa ϭ 7.02 Ϫ 2.11σn (R ϭ 0.904)
(2)
(3)
son with the vinyl group [σp(CH2ϭCH)
ϭ Ϫ0.04;
(4) σpϩ(CH2ϭCH)[15] ϭ Ϫ0.12] shows that the effect of the Cϭ
C bond is very similar, with both PϭC and CϭC having
Clearly, the fit for σpϩ and σn is less satisfactory. Further- only weak interactions with a benzene ring. This feature is
more, Equation (2) furnishes ρ(σp) ϭ 1.49, which, in line in line with the old empirical ‘‘diagonal relationship’’ be-
with expectation, is close to, but slightly smaller than, the tween elements of the second and third periods that pre-
value[26] in pure ethanol (1.57); in contrast, ρ(σpϩ) ϭ 0.88 dicts, inter alia, similarities between carbon and phos-
is smaller and ρ(σn) very much larger than the value in phorus atoms;[31] applied to low-coordinate systems, such
water, a situation that may be considered as additional evi- as (phospha)alkenes YϭC with Y ϭ C or P, this similarity
Eur. J. Org. Chem. 2003, 2049Ϫ2055
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2051

A. Termaten, M. van der Sluis, F. Bickelhaupt
FULL PAPER
Table 2. Chemical shifts δ(31P) of 6 and δ(13Cpara) of substituted benzenes 7 and 8
σn
6
7
8
ϩ
Compound
X
σp
σp
[b][c]
[a][b]
δ(31P)[a]
δ(13Cpara
)
δ(13Cpara
)
a
b
c
d
e
f
g
h
i
j
k
l
m
n
o
NMe2
OH
OMe
Me
SiMe3
H
SnMe3
F
Cl
Br
CHO
COOH
CF3
CN
Ϫ0.83
Ϫ0.37
Ϫ0.27
Ϫ0.17
Ϫ0.07
0.00
0.00
0.06
0.23
0.23
0.42
0.45
0.54
0.66
Ϫ1.70
Ϫ0.92
Ϫ0.78
Ϫ0.31
0.02
Ϫ0.17
Ϫ0.18
Ϫ0.11
Ϫ0.13
Ϫ
0.00
Ϫ
0.06
0.24
0.26
Ϫ
233
244
248
254
262.3
259
260.7
257
116.7
121.1
122.7
127.8
Ϫ
130.6
Ϫ
127.4
129.3
129.8
Ϫ
108.3
113.4
112.6
119.1
Ϫ
122.6
Ϫ
Ϫ
Ϫ
Ϫ
129.4
Ϫ
Ϫ
0.00
Ϫ0.12
Ϫ0.07
0.11
0.15
0.73
0.42
0.61
0.66
0.79
264
264.4
283.1
279.4
Ϫ
0.41
0.53
0.67
0.78
Ϫ
135.4
134.6
136.0
283.9
127.8
129.7
NO2
0.78
284.6[c][d]
[a]
[b]
[c]
[d]
In CDCl3, unless otherwise stated. para-C relative to X; see Scheme 3. In [D6]DMSO. From ref.[24]
with more electron density at the carbon atom adjacent to
the benzene ring, is partly counteracted by a resonance ef-
fect as expressed by 1B. While the carbon atom in 1B and
the phosphorus atom in 1A are both positively charged and
formally both have six electrons, the tricoordinate carbon
atom is energetically less unfavorable than the dicoordinate
phosphorus atom. Moreover, as calculations for the
β-phosphaenones[12] 3 and for their carbon analogue,
acrolein,[32,33] have shown, resonance effects in the π-system
of this type of double bond system tend to be (partially)
compensated by opposite electron shifts in the σ-system.
For the unsubstituted parent phosphaalkene HPϭCH2, ab
initio SCF calculations suggest that its π bond is only
slightly polarized towards the carbon atom; the main con-
tribution to the δϩPϪCδϪ polarization finds its origin in
the σ bond.[34,35]
Figure 2. Hammett plot for values of δ(13Cpara) of 7
In view of this low electronic profile of the PϭC group,
it seems risky to speculate about the difference between our
results obtained for 6 (σp ഠ Ϫ0.1) and those reported for 2
(σp ഠ 0.2).[11,12] Presumably, differences in the structure,
such as the additional aryl group in 2 and the degree of
coplanarity of the aryl groups resulting from steric
crowding,[3Ϫ5,10,11,22] are sufficient to cause the observed
small changes in electronic character.
Finally, having a considerable number of phosphaalkenes
6 available, we investigated also the Hammett-type influence
of the substituents X on values of δ(31P). Previous work on
2 has revealed[3Ϫ5,10,11] that the phosphorus chemical shift
is rather sensitive to substituents at the double bond and
that electron-donating substituents lead to increased shield-
ing of the phosphorus nucleus. Although this observation
seems to be in agreement with a contribution from reson-
ance structure 1B, it may also signal an extreme sensitivity
to small changes resulting from inductive effects.
Figure 3. Hammett plot for values of δ(13Cpara) of 8
has been expressed succinctly in the title of the recent over-
The Hammett relations according to Equations (7), (8),
and (9) were derived from Table 2 and Figure 4. In this case,
view, ‘‘Phosphorus: The Carbon Copy’’.[4]
Apparently, the slight electron-donating character of the the correlations are poorer than in the other Hammett re-
PϭC group that is expected from its overall polarization, lations reported so far; the correlation for σp, however, is
2052
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 2049Ϫ2055

The Substituent Effect of the Phosphaalkenyl Group
FULL PAPER
the best of the three. Moreover, in Equation (7), the refer-
ence value δ(31P) ϭ 259.8 ppm is practically identical to
that of the parent compound 6f [δ(31P) ϭ 259.0 ppm],
whereas Equations (8) and (9) yield δ(31P) ϭ 264.3 and
253.3 ppm, respectively. We feel, therefore, that here, as in
the case of the values of pKa of 6, the correlation with σp
is the more appropriate one.
Scheme 6
One may conclude that, in general, the interaction be-
tween the PϭC bond and the aromatic system is dominated
by inductive effects while resonance effects play a minor
role only, unless a very specific π interaction occurs, as evi-
denced by the value of δ(31P), which may, however, be com-
pensated in part in the σ system.[34,35]
Conclusion
The linear free-energy relationships, derived for the phos-
phaalkenes 6 from the values of pKa of substituted benzoic
acids and from the values of δ(13Cpara) of substituted ben-
zoic acids 7 and bromobenzenes 8, indicate that the (E)-
Mes*PϭCH group is a weak electron donor with mainly an
inductive effect on a neighboring benzene ring; resonance
structures such as 1A and 1B are normally of minor import-
ance. In this regard, it is surprisingly analogous to the Cϭ
C bond, and this result confirms the well-known diagonal
relationship in the Periodic System, justifying the slogan
‘‘Phosphorus: The Carbon Copy’’.[4]
Figure 4. Hammett plot for values of δ(31P) of 6
δ(31P) ϭ 259.81 ϩ 35.62σp (R ϭ 0.941)
δ(31P) ϭ 264.26 ϩ 21.98σpϩ (R ϭ 0.934)
δ(31P) ϭ 253.33 ϩ 46.29σn (R ϭ 0.888)
(7)
(8)
(9)
Experimental Section
Nevertheless, a more detailed analysis of the data for
Equation (9) reveals particularly pronounced deviations for
compounds 6aϪd with the strongly electron-donating sub-
stituents X ϭ NMe2, OH, OMe, and Me; this observation
may be considered as an indication that the purely inductive
constant σn is inadequate in these cases. With appropriate
caution, in view of the limited number of data points (four),
General Remarks: All experiments were performed under N2 in
glassware that had been oven-dried at 150 °C. Solid starting materi-
als were dried under vacuum; liquid starting materials were distilled
under N2. Solvents were distilled from LiAlH4 (pentane, diethyl
ether) or from sodium/benzophenone (THF). NMR spectra were
obtained at room temperature with a Bruker WM 250 spectrometer
(31P: 101.25 MHz) or with a Bruker AC 200 spectrometer (13C:
50.32 MHz; 1H: 200.13 MHz). High-resolution mass spectra
(HRMS) were obtained with a Finnigan MAT 5 spectrometer. El-
emental analyses were performed by Microanalytisches Labor
Pascher, Remagen, Germany. Benzoic acids 7 and bromobenzenes
8 were available from commercial sources. Measurements of the
value of pKa were performed with a Mettler DL25 Titrator by ti-
trating a solution of the benzoic acid (6l, 7) in 85.6% ethanol/water
with 0.1005  aqueous NaOH.
ϩ
we derived Equation (10) with σp for these compounds
and obtained a particularly good correlation (R ϭ 0.992)!
δ(31P) ϭ 258.91 ϩ 15.26σpϩ (R ϭ 0.992)
(10)
This result might be taken as evidence for a resonance
interaction that is stronger than normal, as expressed by
structures 6B (Scheme 6, corresponding to 1B of Scheme 1)
and 6C. For obvious reasons, 6C is not relevant in the case
of electron-withdrawing substituents X. Such an excep-
tional effect is plausible if one assumes that in 6, the value
of δ(31P) is particularly sensitive to perturbations of the π
system, which also has been suggested[29,30] to be a major
[(E)-4-Bromobenzylidene](2,4,6-tri-tert-butylphenyl)phosphane (6j):
Under vigorous stirring at Ϫ65 °C, a solution of n-butyllithium
(1.6  in hexane, 15.7 mL, 25 mmol) was added to a solution of
[36]
Mes*PH2
(6.9 g, 25 mmol) in THF (100 mL). After 5 min, the
solution was warmed to room temperature and stirred for another
15 min. To the red-brown reaction mixture, tert-butyldimethylsilyl
chloride (3.75 g, 25 mmol) was added, followed by stirring for
30 min. n-Butyllithium (1.6  in hexane, 15.7 mL, 25 mmol) was
added to the orange reaction mixture. After 15 min, the reaction
factor responsible for the normally good correlation be-
ϩ
tween values of δ(13C) and σp
.
Eur. J. Org. Chem. 2003, 2049Ϫ2055
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2053

A. Termaten, M. van der Sluis, F. Bickelhaupt
FULL PAPER
mixture was cooled to Ϫ65 °C and p-bromobenzaldehyde (4.61 g,
25 mmol) was added, followed by warming to room temperature
and stirring for 2 h. The solvent was evaporated under reduced
pressure and the residue was dissolved in pentane (100 mL). Tri-
methylsilyl chloride (3.22 mL, 25 mmol) was added to the resulting
solution. After stirring for 30 min, the yellow suspension was fil-
(2,4,6-Tri-tert-butylphenyl)[(E)-4-(trimethylstannyl)benzylidene]-
phosphane (6g): Under vigorous stirring at Ϫ65 °C, n-butyllithium
in hexane (1.6  in hexane, 2.60 mL, 4.0 mmol) was added to a
solution of 6j (1.78 g, 4.0 mmol) in THF (25 mL). After 5 min, tri-
methyltin chloride (1.0 g, 5 mmol) was added. The reaction mixture
was warmed to room temperature and then concentrated to dryness
tered through Al2O3 and the filtrate was concentrated to dryness under vacuum. The residue was extracted with pentane; the extract
to yield practically pure 6j as a yellow powder (10.37 g, 23.3 mmol,
93%); light-yellow crystals (8.41 g, 18.9 mmol, 76%) were obtained
was dried (MgSO4), filtered, and the solvents were evaporated to
dryness. The light-yellow residue (2.10 g, 99%) was recrystallized
after recrystallization from acetonitrile. M.p. 136Ϫ137 °C (no m.p. from acetonitrile to yield 6g as a white powder (1.78 g, 84%). M.p.
1
1
reported in the literature[22]). H NMR (CDCl3): δ ϭ 1.29 (s, 9 H, 140Ϫ142 °C. H NMR (CDCl3): δ ϭ 0.46 [s, 9 H, Sn(CH3)3], 1.53
p-tBu), 1.44 (s, 18 H, o-tBu), 7.35Ϫ7.38 (m, 6 H, Ar-H), 7.96 (d,
(s, 9 H, p-tBu), 1.69 (s, 18 H, o-tBu), 7.60Ϫ7.73 (m, 6 H, Ar-H),
2
2JP,H ϭ 25.4 Hz, 1 H, vinyl-H) ppm. 13C{1H} NMR (CDCl3): δ ϭ
8.30 (d, JP,H ϭ 25.4 Hz, 1 H, vinyl-H) ppm. 13C{1H} NMR
4
31.4 [s, p-C(CH3)3], 33.8 [d, JP,C ϭ 7.0 Hz, o-C(CH3)3], 35.0 [s, p-
(CDCl3): δ ϭ Ϫ9.6 [s, Sn(CH3)3], 31.4 [s, p-C(CH3)3], 33.8 [d,
C(CH3)3], 38.2 [s, o-C(CH3)3], 121.6 [d, 5JP,C ϭ 9.3 Hz, p-C6H4(C)],
4JP,C ϭ 7.0 Hz, o-C(CH3)3], 34.9 [s, p-C(CH3)3], 38.2 [s, o-C(CH3)3],
3
3
121.8 [s, m-Ar(C)], 127.1 [d, JP,C ϭ 22.1 Hz, o-C6H4(C)], 131.8 [d, 121.7 [s, m-Ar(C)], 125.2 [d, JP,C ϭ 22.0 Hz, o-C6H4(C)], 136.2 [d,
1
1
4JP,C ϭ 3.0 Hz, m-C6H4(C)], 138.5 [d, JP,C ϭ 53.3 Hz, i-Ar(C)], 4JP,C ϭ 2.6 Hz, m-C6H4(C)], 139.0 [d, JP,C ϭ 53.7 Hz, i-Ar(C)],
2
2
138.9 [d, JP,C ϭ 14.1 Hz, i-C6H4(C)], 149.8 [s, p-Ar(C)], 154.0 [s, 140.0 [d, JP,C ϭ 13.8 Hz, i-C6H4(C)], 142.7 [s, p-C6H4(C)], 149.6
1
1
o-Ar(C)], 174.4 (d, JP,C ϭ 34.6 Hz, PϭC) ppm. 31P NMR
[s, p-Ar(C)], 154.0 [s, o-Ar(C)], 175.8 (d, JP,C ϭ 34.6 Hz, PϭC)
(CDCl3): δ ϭ 264.4 ppm. C25H3479BrP: calcd. 444.1582, found
ppm. 31P NMR (CDCl3): δ ϭ 260.7 ppm. C28H43P116Sn calcd.
526.2123, found (HRMS EI) 526.2126. C28H43PSn (529.30): calcd.
C 63.53, H 8.19; found C 64.82, H 8.30.
(HRMS EI) 444.1585.
[(E)-4-Cyanobenzylidene](2,4,6-tri-tert-butylphenyl)phosphane (6n):
According to the procedure described for 6j, Mes*PH2 (1.1 g,
4 mmol) was converted into 6n by reaction with p-cyanobenzal-
dehyde (0.52 g, 4 mmol) to give a light-yellow powder of 6n (1.49 g,
3.81 mmol, 95%); light-yellow needles (1.14 g; 2.92 mmol; 73%)
were obtained after recrystallization from acetonitrile. M.p.
4-{[(E)-(2,4,6-Tri-tert-butylphenyl)phosphanylidene]methyl}benz-
aldehyde (6k): Under vigorous stirring at Ϫ65 °C, n-butyllithium in
hexane (1.6  in hexane, 2.60 mL, 4.0 mmol) was added to a solu-
tion of 6j (1.78 g, 4.0 mmol) in THF (25 mL). After 5 min, N,N-
dimethylformamide (0.31 mL, 4 mmol) was added to the dark-
orange reaction mixture. The reaction mixture turned green and
was warmed to room temperature. After quenching with aqueous
NH4Cl, the solvents were evaporated to dryness under vacuum.
The yellow residue was extracted with pentane (25 mL); the extract
was dried (MgSO4), filtered, and the solvents were evaporated to
dryness. The light-yellow residue (1.47 g, 93%) was recrystallized
from acetonitrile to yield 6k as yellow needles (1.30 g, 82%). M.p.
1
98Ϫ101 °C. H NMR (CDCl3): δ ϭ 1.38 (s, 9 H, p-tBu), 1.52 (s,
2
18 H, o-tBu), 7.47Ϫ7.62 (m, 6 H, Ar-H), 8.07 (d, JP,H ϭ 25.1 Hz,
1 H, vinyl-H) ppm. 13C{1H} NMR (CDCl3): δ ϭ 31.2 [s, p-
4
C(CH3)3], 33.8 [d, JP,C ϭ 6.9 Hz, o-C(CH3)3], 34.9 [s, p-C(CH3)3],
5
38.1 [s, o-C(CH3)3], 110.4 [d, JP,C ϭ 8.2 Hz, p-C6H4(C)], 119.1 (s,
3
CN), 121.9 [s, m-Ar(C)], 125.9 [d, JP,C ϭ 22.4 Hz, o-C6H4(C)],
4
1
132.4 [d, JP,C ϭ 3.0 Hz, m-C6H4(C)], 137.6 [d, JP,C ϭ 53.1 Hz, i-
1
125Ϫ127 °C. H NMR (CDCl3): δ ϭ 1.28 (s, 9 H, p-tBu), 1.43 (s,
2
Ar(C)], 144.0 [d, JP,C ϭ 14.6 Hz, i-C6H4(C)], 150.1 [s, p-Ar(C)],
2
18 H, o-tBu), 7.16Ϫ7.75 (m, 6 H, Ar-H), 8.25 (d, JP,H ϭ 25.1 Hz,
153.9 [d, 2JP,C ϭ 1.5 Hz, o-C6H4(C)], 172.5 (d, 1JP,C ϭ 36.0 Hz, Pϭ
C) ppm. 31P NMR (CDCl3): δ ϭ 283.9 ppm. C26H34NP: calcd.
391.2429, found (HRMS EI) 391.2428. C26H34NP (391.51): calcd.
C 79.75, H 8.76; found C 79.78, H 8.77.
1 H, vinyl-H), 9.88 (s, 1H CHO) ppm. 13C{1H} NMR (CDCl3):
4
δ ϭ 31.2 [s, p-C(CH3)3], 33.7 [d, JP,C ϭ 7.0 Hz, o-C(CH3)3], 34.9
[s, p-C(CH3)3], 38.1 [s, o-C(CH3)3], 121.9 [s, m-Ar(C)], 126.0 [d,
3JP,C ϭ 22.4 Hz, o-C6H4(C)], 130.2 [d, 4JP,C ϭ 2.9 Hz, m-C6H4(C)],
5
1
135.2 [d, JP,C ϭ 7.1 Hz, p-C6H4(C)], 138.0 [d, JP,C ϭ 53.4 Hz, i-
2
(2,4,6-Tri-tert-butylphenyl)[(E)-4-(trimethylsilyl)benzylidene]phos-
phane (6e): Under vigorous stirring at Ϫ65 °C, n-butyllithium in
hexane (1.6  in hexane, 1.98 mL, 3.0 mmol) was added to a solu-
tion of 6j (1.32 g, 3.0 mmol) in THF (20 mL). After 5 min, tri-
methylsilyl chloride (0.36 mL, 3.0 mmol) was added. The reaction
mixture was warmed to room temperature and then concentrated
to dryness under vacuum. The residue was extracted with pentane;
the extract was dried (MgSO4), filtered, and the solvents were eva-
porated to dryness. The light-yellow residue (1.26 g, 96%) was
recrystallized from acetonitrile to yield 6e as light-greenish needles
Ar(C)], 145.7 [d, JP,C ϭ 14.0 Hz, i-C6H4(C)], 150.0 [s, p-Ar(C)],
153.9 [s, o-Ar(C)], 173.3 (d, 1JP,C ϭ 36.0 Hz, PϭC), 191.4 (s, CHO)
ppm. 31P NMR (CDCl3): δ ϭ 283.1 ppm. C26H35OP: calcd.
394.2426, found (HRMS EI) 394.2427. C26H35OP (394.51): calcd.
C 79.15, H 8.95; found C 79.31, H 8.89.
4-{[(E)-(2,4,6-Tri-tert-butylphenyl)phosphanilidene]methyl}benzoic
Acid (6l): Under vigorous stirring at Ϫ65 °C, n-butyllithium in hex-
ane (1.6  in hexane, 2.60 mL, 4.0 mmol) was added to a solution
of 6j (1.78 g, 4.0 mmol) in THF (25 mL). After 5 min, an excess of
(1.05 g, 80%). M.p. 181Ϫ183 °C. 1H NMR (CDCl3): δ ϭ 0.30 [s, 9 CO2 was sublimed into the dark-orange reaction mixture, which
H, Si(CH3)3], 1.39 (s, 9 H, p-tBu), 1.55 (s, 18 H, o-tBu), 7.48Ϫ7.62 was then stirred for another 5 min and warmed to room tempera-
2
(m, 6 H, Aryl-H), 8.17 (d, JP,H ϭ 25.5 Hz, 1 H, vinyl-H) ppm. ture. Trimethylsilyl chloride (1.2 mL, 9.5 mmol) was added to the
13C{1H} NMR (CDCl3): δ ϭ Ϫ1.2 [s, Si(CH3)3], 31.4 [s, p- resulting yellow suspension; the precipitate dissolved, water
4
C(CH3)3], 33.8 [d, JP,C ϭ 7.0 Hz, o-C(CH3)3], 34.9 [s, p-C(CH3)3], (0.5 mL, 28 mmol) was added, and the mixture was concentrated
3
38.2 [s, o-C(CH3)3], 121.7 [s, m-Ar(C)], 124.9 [d, JP,C ϭ 21.9 Hz, to dryness under vacuum. The residue was extracted with CHCl3
o-C6H4(C)], 133.7 [d, 4JP,C ϭ 2.7 Hz, m-C6H4(C)], 139.0 [d, 1JP,C ϭ
(20 mL); the extract was washed with water, dried (MgSO4), fil-
tered, and then the solvents were evaporated to dryness. The yellow
residue (1.59 g, 97%) was recrystallized from acetonitrile to yield 6l
2
53.6 Hz, i-Ar(C)], 140.4 [s, p-C6H4(C)], 140.4 [d, JP,C ϭ 13.9 Hz,
1
i-C6H4(C)], 149.6 [s, p-Ar(C)], 154.0 [s, o-Ar(C)], 175.6 (d, JP,C
ϭ
34.8 Hz, PϭC) ppm. 31P NMR (CDCl3):
δ ϭ
262.3 ppm. as yellow needles (1.09 g, 66%). M.p. 215Ϫ217 °C. 1H NMR
C28H43P28Si: calcd. 438.2872, found (HRMS EI) 438.2872. (CDCl3): δ ϭ 1.29 (s, 9 H, p-tBu), 1.44 (s, 18 H, o-tBu), 7.38 (s, 2
3
4
C28H43PSi (438.68): calcd. C 76.65, H 9.89; found C 76.79, H 9.85.
H, Ar-H), 7.56 (dd, JH,H ϭ 8.1, JP,H ϭ 3.1 Hz, 2 H, o-C6H4-H),
2054
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 2049Ϫ2055

The Substituent Effect of the Phosphaalkenyl Group
FULL PAPER
[15]
[16]
7.97 (d, 3JH,H ϭ 8.1 Hz, 2 H, m-C6H4-H), 8.10 (d, 2JP,H ϭ 23.5 Hz,
1 H, vinyl-H), 11.97 (br. s, 1 H, COOH) ppm. 13C{1H} NMR
C. Hansch, A. Leo, R. W. Taft, Chem. Rev. 1991, 91, 165Ϫ195.
J. Shorter, in: Similarity Models in Organic Chemistry, Biochem-
istry and Related Fields (Eds.: R. I. Zalewski, T. M. Krygowski,
J. Shorter), Elsevier, Amsterdam, 1991, chapter 2.
A. J. Hoefnagel, B. M. Wepster, J. Chem. Soc., Perkin Trans. 2.
1989, 977Ϫ986.
A. J. Hoefnagel, B. M. Wepster, Collect. Czech. Chem. Com-
mun. 1990, 55, 119Ϫ135.
H. van Bekkum, P. E. Verkade, B. M. Wepster, Recl. Trav.
Chim. Pays-Bas 1959, 78, 815Ϫ850.
M. Yoshifuji, K. Toyota, I. Matsuda, T. Niitsu, N. Inamoto,
K. Hirotsu, T. Higuchi, Tetrahedron 1985, 44, 1363Ϫ1367.
K. Issleib, H. Schmidt, E. Leißring, Z. Chem. 1986, 26,
406Ϫ407.
V. D. Romanenko, A. V. Ruban, A. N. Chernega, M. I. Povol-
otskii, M. Yu. Antipin, Yu. T. Struchkov, L. N. Markovski, J.
Gen. Chem. USSR 1989, 59, 1528Ϫ1534.
M. van der Sluis, A. Klootwijk, J. B. M. Wit, F. Bickelhaupt,
N. Veldman, A. L. Spek, P. W. Jolly, J. Organomet. Chem. 1997,
529, 107Ϫ119.
S. Shah, J. D. Protasiewicz, Chem. Commun. 1998, 1585Ϫ1586.
CRC Handbook of Chemistry and Physics (Ed.: D. R. Lide),
CRC Press, Boca Raton, 1998, section 8, p. 52.
P. R. Wells, Linear Free-Energy Relationships, Academic Press,
New York, 1968, pp. 12Ϫ13.
G. L. Nelson, G. C. Levy, J. D. Cargioli, J. Am. Chem. Soc.
1972, 94, 3089Ϫ3094.
J. Bromilov, R. T. Brownlee, V. O. Lopez, R. W. Taft, J. Org.
Chem. 1979, 44, 4466Ϫ4470.
J. Bromilov, R. T. Brownlee, D. J. Craik, M. Sadek, R. W. Taft,
J. Org. Chem. 1980, 45, 2429Ϫ2438.
J. Bromilov, R. T. Brownlee, D. J. Craik, M. Sadek, Magn. Re-
son. Chem. 1986, 24, 862Ϫ871.
J. P. Huheey, Inorganic Chemistry, 3rd ed., Harper International
SI Edition, Cambridge, 1983, pp. 822Ϫ824.
K. B. Wiberg, R. E. Rosenberg, P. R. Rablen, J. Am. Chem.
Soc. 1991, 113, 2890Ϫ2898.
K. B. Wiberg, P. R. Rablen, M. Marquez, J. Am. Chem. Soc.
1992, 114, 8654Ϫ8668.
W. W. Schoeller, J. Chem. Soc., Chem. Commun. 1985,
4
(CDCl3): δ ϭ 31.2 [s, p-C(CH3)3], 33.7 [d, JP,C ϭ 6.9 Hz, o-
C(CH3)3], 34.9 [s, p-C(CH3)3], 38.1 [s, o-C(CH3)3], 121.8 [s, m-
[17]
[18]
[19]
[20]
[21]
[22]
3
5
Ar(C)], 125.4 [d, JP,C ϭ 22.0 Hz, o-C6H4(C)], 127.9 [d, JP,C
ϭ
7.3 Hz, p-C6H4(C)], 130.6 [d, 4JP,C ϭ 2.4 Hz, m-C6H4(C)], 138.3 [d,
2
1JP,C ϭ 53.4 Hz, i-Ar(C)], 145.0 [d, JP,C ϭ 14.0 Hz, i-C6H4(C)],
149.9 [s, p-Ar(C)], 153.9 [s, o-Ar(C)], 171.8 (s, COOH), 173.5 (d,
1JP,C ϭ 35.7 Hz, PϭC) ppm. 31P NMR (CDCl3): δ ϭ 279.4 ppm.
C26H35O2P: calcd. 410.2375, found (HRMS EI) 410.2376.
C26H35O2P (410.51): calcd. C 76.06, H 8.60; found C 76.30, H 8.78.
Acknowledgments
We thank a referee for valuable suggestions concerning several
aspects of the Hammett relation.
[23]
[1]
[24]
[25]
K. S. Pitzer, J. Am. Chem. Soc. 1948, 70, 2140Ϫ2145.
[2]
R. S. Mulliken, J. Am. Chem. Soc. 1950, 72, 4493Ϫ4503.
[3]
Multiple Bonds and Low Coordination in Phosphorus Chemistry
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
(Eds.: M. Regitz, O. J. Scherer), Thieme, Stuttgart, 1990.
Phosphorus: The Carbon Copy (Eds.: K. B. Dillon, F. Mathey,
[4]
J. F. Nixon), Wiley, Chichester, 1998.
D. E. C. Corbridge, Phosphorus 2000, Chemistry, Biochemis-
[5]
try & Technology, Elsevier, Amsterdam, 2000, chapter 6.
L. Pauling, The Nature of the Chemical Bond, 3rd. ed., Cornell
[6]
University Press, Ithaca, NY, 1960, pp. 88Ϫ93.
[7]
A. L. Allred, E. G. Rochow, J. Inorg. Nucl. Chem. 1958, 5, 264.
[8]
G. Simons, M. E. Zandler, E. R. Talaty, J. Am. Chem. Soc.
1976, 98, 7869Ϫ7870.
[9]
R. J. Boyd, S. L. Boyd, J. Am. Chem. Soc. 1992, 114,
1652Ϫ1655.
[10]
T. van der Does, F. Bickelhaupt, Phosphorus, Sulfur Silicon
1987, 30, 515Ϫ518.
T. van der Does, F. Bickelhaupt, Phosphorus, Sulfur Silicon
[11]
1990, 49Ϫ50, 281Ϫ284.
M. van der Sluis, F. Bickelhaupt, N. Veldman, H. Kooijman, A.
[12]
L. Spek, W. Eisfeld, M. Regitz, Chem. Ber. 1995, 128, 465Ϫ476.
L. P. Hammett, Physical Organic Chemistry, McGraw-Hill,
334Ϫ335.
[13]
A. B. Rozhenko, W. W. Schoeller, M. I. Povolotskii, Magn. Re-
son. Chem. 1999, 37, 551Ϫ563.
New York, 1940, chapter 7.
D. F. Ewing, in: Correlation Analysis in Chemistry (Eds.: N. B.
[14]
A. H. Cowley, J. E. Kilduff, T. H. Newman, M. Pakulski, J.
Am. Chem. Soc. 1982, 104, 5820Ϫ5821.
Chapman, J. Shorter), Plenum Press, New York, 1978, chap-
ter 8.
Received November 7, 2002
Eur. J. Org. Chem. 2003, 2049Ϫ2055
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2055