close

Вход

Забыли?

вход по аккаунту

?

Nitrile Imines From Matrix Characterization to Stable Compounds.

код для вставкиСкачать
REVIEWS
Nitrile Imines : From Matrix Characterization to Stable Compounds
Guy Bertrand* and Curt Wentrup"
The concept of 1,3-dipolar cycloaddition was developed, starting in the late
1950s. largely by Rolf Huisgen and his
students in Munich, and it has led to one
of the most versatile methods for the
construction of five-membered ring heterocycles. Although first known only as
transient intermediates, nitrile imines
have been at the heart of mechanistic
studies of this type of cycloaddition
reactions. Hundreds of mechanistic
papers appeared in 1960s and 1970s;
reliable spectroscopic observations were
achieved in the early 1980s both at low
temperatures and in the gas phase; finally, the first crystalline nitrile imine was
reported in 1988. The unusual structures
found by X-ray analyses as well as the
1. Introduction
Until recently, nitrile imines l,[l-'] like several other 1,3dipoles, were considered reactive intermediates. Since their discovery by Huisgen at al. in 1959,['"] nitrile imines have been
generated in situ and used in numerous 3,3-dipolar cycloaddition reactions leading to pyrazoles. pyrazolines, and other heterocyclic compounds.[6$ Furthermore, nitrile iniines have been
at the heart of mechanistic studies of the 1,3-dipolar cycloaddition reaction and of the controversy over Firestone's diradical
pathway."]
Nevertheless. because of their character as reactive intermediates, there has been very little direct experimental evidence on
the structures of nitrile imines. The most important canonical
structures['] are indicated in formulas 1 a-f (Scheme 1 ) . These
[*I
Dr. G. Bertrand
Laboratoire de Chiinie de Coordination d u CNRS
205. route de Narbonne, 31077 Toulouse Cedex (France)
Telefax. Int. code + (61j55-3003
Prof. Dr. C. Wentrup
Department of Chemistry
The Univeryity of Queensland
Brisbane, Old. 4072 (Australia)
Telefax' Int code +73654580
[**I
In thia rmiebv formula numbers with the letter n refer t o nitrile imines. and
those with the letter d refer to diazomethane derivatives.
facile rearrangements observed experimentally have fostered a new interplay
between experiment and theory. The
story of nitrile imines, from matrix characterization to stable compounds, nicely
illustrates the role that main group elements can play in organic chemistry.
d o not say anything about the actual structures, not even
whether the nitrile imines are linear or bent. Ab initio calculations by Houk et
l o ] reveal that the parent compound formonitrile imine 2n[**] may exist in either a planar (propargylic.
2a) o r a nonplanar, bent (allenic, 2b) form; 2b is favored by
2.2 kcal mol-' when the STO-3G basic set is employed, but the
progargylic form 2 a is favored by 3.9 kcalmol-' at the 4-31G
level. Thus, the unsubstituted
nitrile imine is described as a
+ N
H - +
"floppy" molecule,'91 which
H-CEN' 'H
'C=N&
N*H
may be able to adapt its ge2a
2b
ometry easily in response
to substituents. Similar ~ a l c u l a t i o n s ~for
~ ] nitrile ylides
+ (R-C-N-CR,)
indicate that these prefer bent, nonplanar
+
structures, whereas in nitrile oxides ( R - C s N - O - )
the linear
form is theoretically favored, in agreement with experiment. For
both nitrile imines and nitrile ylides, electron-withdrawing substituents. particularly at the N o r CR, termini, are expected to
promote the planar, propargylic forms.[', 'I
The situation has changed recently. Nitrile imines have become accessible as stable, sometimes crystalline solids or distillable liquids. Accordingly, their structures and properties can
now be probed in a much more direct manner. It has been
found, as described in Section 5.1, that all stable nitrile imines
examined by X-ray crystallography to date definitely possess
bent, helical backbones, more of the "allenic" type 2 b. Furthermore, high-level calculations"21 also predict that the bent, helical structure represents the lowest energy form of nitrile imines
(see Section 5.1 ) .
Another problem concerns the rearrangements of nitrile
imines and of 1,3-dipoles in general. Rearrangements to diazo
compounds (R,C=N=N-) have been postulated several times
G. Bertrand and C. Wentrup
REVIEWS
(see Section 7). If this is a concerted process, it constitutes a
highly interesting and formally “forbidden” 1,3-shift. Likewise,
both thermal and photochemical rearrangements to carbodiimides ( R N = C = N R ) have been reported (Section 7). With the
emergence of stable, isolable nitrile imines, it has become possible to probe such rearrangements directly.
R-CCl=N-NHR’
2. History
+
-
MeLi
--+
2O0C
-CH4
OH-
R-CSN-PN-R’
+
C1- + H20
(2)
4
The parent formonitrile imine 2 n is still unknown. It was once
thought to be formed on deprotonation of diazomethane with
methyllithium at room temperature, followed by hydrolysis
[Eq. (l)].’’31 However, 34 years later, it was established, largely
H$=N=N
these compounds generate nitrile imines, which undergo cycloaddition reactions to furnish pyrazoles. The possibility of
formation of disubstituted isodiazomethanes (“isodiazometani
bisostituiti”) was mentioned in 1946 and 1948 by Fusco and
Romani, who also considered other isomers of diazomethane
[Eq. (2)].[201
H t = i = i Li+
H20
--b
+ -
[.-Cdi-N-H]
2n
--+
- +
CeN-NH2 (1)
Even earlier, a compound had been postulated which would
nowadays be called diphenylnitrile imine. In 1902 Bamberger
and Grob treated x-nitrobenzylidenephenylhydrazine( 5 ) with
sodium methoxide in methanol and obtained tetraphenyldihydrotetrazine (6) in 39% yield [Eq. ( 3 ) ].[”, 221 This was explained
3
Ph-C=N-NH-Ph
on the basis of the N M R and infrared spectra, that this “isodiazomethane” was in fact isocyanoamine 3.[14-161
It is not
known whether nitrile imine 2 n is actually an intermediate in
this reaction.“ ’I Isocyanoamine 3 is rapidly converted to diazomethane under the influence of alkali metal hydroxides. It is
interesting to note that E. Miiller called 3 “Kiiallsaureaiiiid”,
that is, fulminic acid amide. However, W. Beck later established
that fulminic acid has the structure HCNO, not H O N C as had
been believed until then.[’81H,N-NC (3) may thus be considered as an amide of HONC (isofulminic acid) but not of fulminic
acid.
In the 2930s and 1940s Fusco and his students in Italy carried
out work on hydrazonoyl chlorides 4.[19]It is now known that
NO2
NaOMe
___)
neoH
5
Ph
Ph</
N-N/
,&Ph
(3)
ph/N-N
6
with the words: “Das Natriumsalz des Isonitrokorpers zerfiillt
in Nitrit und den Molekularrest C6H,-?-N=N-C6H,,
welcher sofort zum Tetrazolinderivat polymerisiert”.[’ b1 As
Huisgen et al. later pointed out,[221this “molecular residue”
corresponds to the canonical structure 1 e of a nitrile imine.
Nitro compounds such as 5 are versatile precursors of nitrile
imines, in particular in cases where the corresponding hydrazonoyl halides are inaccessible. It is also known that the “nitrohydrazone route” gives particularly high yields of dihydrotetrazines: this cannot be due to direct dimerization of the nitrile
imines but rather to a reaction between one molecule of nitrile
Guy Bertrand was born in Limoges, France, on July17, 1952. He graduated (inglnieur ENS C M ) from the University of Montpellier and moved to Toulouse as an “Attache de Recherche
CNRS” in 1975. He defended his “ThPse d’Etat” at the University of Toulouse in 1979. After
one year at Sanqfi Recherche Company, he came back to the University Paul Sabatier as a
“Charge de Recherche”. In 1988 he went to the Laboratoire de Chiniie de Coordination du
C N R S where he is non’ “Directeur de Recherche CNRS”. B,y employing main group elements,
his group is trying to isolate at room temperature species hitherto believed to be only transient
intermediates. They use these species in organic synthesis, as ligands for transition metal
comple.ues, and as precursors for new materials.
Curt Wentrup M’US born in Holtug, Denmark, on July 11, 1942. He receivedhis “Diplom (cand.
scient.)” in chemistry in 1966,from the University of Copenhagen for work with K. A . Jensen
on alkyl cyanates, his P1z.D. in 1969 from the Australian National University. Canberra, and
his Dr. scient. in 1977,from the University of Copenhagen for work on carbene and nitrene
rearrangements. After a period as a postdoctoral associate and later “Privutdozent with Hans
Dahn at the University ?f Lausanne, Switzerland (1969-1976), he joined the chemistry faculty
at the University of Marburg, Germanj (1976-1985). h 1985 he returned to Australia as
projessor of organic chemistry and head of the Organic Chemistry Section, Department qf
Chemistry, University qf Queensland, Brisbane. His re.warch interests include reactive intermediates and unusual molecules qften generated hj’,jla.rlzvacuum pyrolysis and studied by matri-x
isolation techniques.
”
528
Angeir Chem Int Ed EnRl 1994, 33, 521 545
REVIEWS
Nitrile Imines
imine with the long-lived anion of the starting material 5.[221
In
general, it appears that the formation of dihydrotetrazines in
substantial yields is indicative of reaction between the nitrile
imine and its precursor.[22-2s1
The true dimerization of nitrile imines in the absence of other
reaction partners leads to bis(azo)ethylenes of the type 7, which
can be isolated in certain cases[26-281[Eq. (4)] (see also Section 3.2).
N
N-N, R'
'
R -$=N-NH-R
R -C=N-fi-R '.ll+
x 9
NoZ 16
X:Cl.Br
bas\
1.'-hv
+ -
R-C EN-N- R '
T
p b ( o * c y
R -C=N-NH-R'
17
yv
hv
f
N-N,
0
N
'c-01+1
+ -
2 R-CZN-N-R'
7
R"
0
I
R - I16 's=O
N-N,
R'
20
3. Nitrile Imines as Reactive Intermediates
19
R'
N-C,
(4)
._+
o,,
i
R-Cn C=O
R 18
f 0+o
Ph'
R -$=N-fi-COR
C-N,
R
21
22
Scheme 3 . Methods used for the generation of transient nitrile mines.
3.1. Generation
Since very authoritative reviews of the classical chemistry of
nitrile imines and other 1,3-dipoles exist,[Z-7Jonly a brief sketch
will be provided here. In the first deliberate generation and
trapping of nitrile imines in 1959,L'a1
both the thermal decomposition of 2,5-disubstituted tetrazoles 8 at 150-200°C and the
base-induced dehydrochlorination of the hydrazonoyl chloride 9 were examined. Diphenylnitrile imine 1 I n was trapped in
situ in nucleophilic additions of aniline and phenol (giving compounds 10 and 12, respectively) and in 1,3-dipolar cycloaddition
reactions with benzaldehyde, benzonitrile, and dicyclopentadiene to produce 13, 14 and 15, respectively (Scheme 2).
photolysis of 19, actually generates nitrile imines is disputed.'261
Although the same products and the same orientations of cycloaddition are obtained with 19 (R = R' = Ph) as with
diphenylnitrile imine (11n) generated from 2,5-diphenyltetrazole (8b),[32a1the photolysis of 19 in the absence of a dipolarophile was found to be extremely sluggish, and it was postulated
that CO and benzoylazobenzene 23 formed rather than the nitrile imine (Scheme 4).[261
,.
Scheme 4. Photolysis of 1,3.4oxadiazol-S(4 H)-ones 19.
N
P h - 0
N-N
0
Ph-C C=O
I1
I
N-N,
19
Ph
3
Ph-CsN;N;PPh
+ -
-co
Ph-f-N=N-Ph
0 23
Ph-Y-N-NH-R
c1 9
Ph
llrd. R = M e
I 0
15
(75%)
PhOH/
Ph-C=N-NH-R
OPh
12
(79%)
0
Ph-$
N-N,
+'Hh
13 R
(75%)
Ph-i't-Ph
N-N,
R
14
Thus, the cycloadditions starting with 19, which mimic nitrile
imine cycloadditions, may involve direct reaction between an
excited state of 19 and the dipolarophile. Interestingly, it was
reported that no nitrile imine dimers were formed in the photolysis of 19.L32a1
Although the yields of dimers are often low, their
total absence may be a telltale sign that no nitrile imine was
formed.
Less common precursors for nitrile imines are 1,2,3,4oxathiadiazole-2-oxides 2O,[j31 1,2,4-oxadiazol-5(4H)-ones
21 ,[341 and pyridinium betaines 22.[35'
(63%)
Scheme 2. First generation and trapping of a transient nitrile imine. Herc and in
subsequent qchemes the percentages in parentheses refer to the yields.
3.2. Methods for the Direct Observation of Transient
Nitrile Imines
The generation of nitrile imines by photolytic decomposition
of tetrazoles 8 was reported soon afterwards." b1 These methods,
as well as the use of salts of aci-nitro-azo compounds (r-nitrohydrazones) 16,122.
291 have become standard procedures for nitrile
imine production. In addition, the oxidation of hydrazones 17
with lead tetraa~etate,[~'Ithe photolysis of sydnones
'I1
and the photolysis of 1,3,4-oxadiazol-5(4 H)-ones 19[321also
provide nitrile imines (Scheme 3). Whether the latter reaction,
Surprisingly. before 1988, there was only one report on the
direct observation of a nitrile imine in solution. Grundmann
and F l ~ r y [used
~ ~ ]the thermal decomposition of the aci-nitroazo compound 24 to prepare dimesitylnitrile imine 25n in solvents such as THF and benzene (Scheme 5 ) . Potassium nitrite
formed quantitatively in 10-20 minutes, and the solution
turned orange, presumably because of the formation of 25 n.
When the reaction was carried out in the presence of dimethyl
fumarate, cycloadduct 26 was formed in over 90% yield. When
An,qrw. Chem 1171. E d Engl.
1994. 33, 527-545
529
G. Bertrand and C. Wentrup
REVIEWS
+ -
70°C
\f
-02.,K+
24
:*
H,
N=N- Ar
Ar
Ar-CsN-N-Ar
25n
THF
15 min
+H
F. .H
Ar-5 C <
b
I R
N-N,
Ar
26
-KN02
Nitrile imine 33n was generated analogously by photolysis of
tetrazole 32 and observed by UV spectroscopy (Scheme 7) .[281
Ph
Scheme 5 . Gcneration and trapping of dimesitylnitrile imine 25n. Ar
methylpheiiyl (mesityl): R = C0,Me.
=
2.4.64"-
/
h V , EPA
-1
the reaction was repeated in the absence of the trapping agent
and the solution filtered to remove KNO, before dimethyl fumarate was added, 26 was still isolated in 18% yield. The I R
spectrum of the orange solution featured a band at 2200 cm ',
which disappeared in the course of 24 h. In the spectra of reaction mixtures containing dimethyl fumarate, the same band disappeared in 1 -2 h (surprisingly slowly). This experiment constitutes the first direct observation of a nitrile imine.
2.5-Disubstituted tetrazoles 8 have been used to generate nitrile imines both photochemically and thermally (flash vacuum
pyrolysis) for direct UV and TR spectroscopic observation.
Heimgartner et al.r361photolyzed 2,5-diphenyltetrazole (8b) at
77 K in organic glasses (etherlpeiitanei'ethanol (EPA), 2methylpentane (2MP). or 2,3-dimethylbutane-pentane (DMBP)) .
Upon irradiation with light of wavelength 280-290 nm, a new
species was observed (A,,, = 378 nm) which, on the basis of its
behavior, was identified as diphenylnitrile imine 11 n (Scheme 6).
E tCOOH
-
hV
EPA
Ph-<*y
LPh-CaN-N-Ph
N-N,
77
lln
8b
Ph
+
E t O H /EPA
,/130K
Ph-FN-NH-Ph
O E t 29
'")(
N=N-Ph
Ph
N=N-Ph
L
>130K
RT
Ph
6
ph-$*y
N-N,
28
Ph
27
DtlBAEtCOOH
135K\
P
200 K
Ph-CZN-NH-Ph d Ph-C-NH-N-C-E t
b C E t 30
8
8 31
0
Scheme 6. Synthesis and characterization of diphenylnitrile imine I 1 n. RT
temperature.
=
room
Scheme 7 Photolysis of tetrazole 32
Warming nitrile imine 33n to room temperature resulted in an
intramolecular[3811,3-dipolar cycloaddition reaction giving 34.
This compound was also obtained in 84 YOyield in the preparative photolysis of 32 at room temperature. Preparative photolysis in the presence of a 1000-fold excess of propionic acid gave
hydrazide 36 in 83% isolated yield, presumably via an initial
adduct 35. Thus, carboxylic acids were found to be extremely
efficient traps for nitrile imines, and the corresponding reactions
compete with the intramolecular 1,3-dipolar cycloaddition reaction (Scheme 7) .[281
The direct detection of thermally generated nitrile imines is
more difficult because of rapid rearrangement reaction^[^^-^^]
(see Section 7). However, N-trimethylsilyl-C-phenylnitrile
imine (37n) turned out to be quite stable under flash vacuum
pyrolysis (FVP) conditions (Scheme 8). It was directly observed
N
ph-(+y
N-N.
710 K
-b
10-3 Torr
H
37n
Sine3
HCzCR
R : Me3SiOC0
R': MeOCO
C
Ph5''iR'
N-N,
39
Sineg
Scheme 8. Thermal generation and trapping of nitrile imme 370
Toubro and
carried out similar photolyses
(A = 250 nm) of 8 b in EPA o r poly(viny1 chloride) (PVC) at
85 K and observed the UV spectrum of 11 n at 377 nm. Moreover. they observed a major IR band at 2228 cm- (PVC, 85 K ) ,
which was ascribed to the C = N stretch in 11 n. This assignment
was corroborated by generation of "N isotopomers by photolysis of suitably labeled tetrazoles. Nitrile imine 11 n is stable in
the glass at temperatures below roughly 130 K . In the absence
of trapping agents, it dimerizes above 130 K, in all likelihood
forming bis(azo)ethylene 27. The latter compound decomposes
at room temperature. but 1,2,3-triazole 28 is not formed in the
process (28 is formed on room-temperature photolysis of
8b).[361
Nitrile imine 11 n is trapped by ethanol above the softening
point of the EPA matrix, giving hydrazone 29 at 130 K.[281With
propionic acid in DMBP glass at 135 K, 11 n was trapped to give
initially 30, which on warming to 200 K isomerized to hydrazide
31 (Scheme 6). The reaction 30 -+ 31 was monitored by UV
'
530
in the gas phase by mass spectrometry and distinguished from
the isomeric carbodiimide and diazo compound by collisional
activation mass spectrometry.1431Deposition of 37n at 77 K
permitted the recording of the IR spectrum, which featured a
strong absorption at 2230cm-'. The IR spectrum was also
distinct from those of the isomeric diazo compound and carbodiimide.
Nitrile imine 37n was trapped in high yields in 1.3-dipolar
cycloadditions in the gas phase. More importantly, it could also
be cocondensed with the trapping reagent at 77 K and observed
by IR spectroscopy (absorption band at 2230 cm-I). Warming
to 170 K resulted in cycloaddition reaction. giving the products
38 and 39 (Scheme 8) in almost quantitative yields. Trapping
37 n with methyl propiolate gave 39 regiospecifically ; the
product was identical to the compound obtained on silylation of
methyl 5-phenylpyrazole-3-carboxylate,and the latter compound was also obtained on hydrolysis of 39 in 90% ethanol.
Angebt. Chem.
In!. Ed En,@ 1994. 33. 527-545
REVIEWS
Nitrile Imines
This work made it possible also to obtain the first photoelectron spectra of nitrile imines by directly monitoring the FVP reactions. The first ionization potential was
t
-
7.85 eV for Ph-CEN-N-SiMe,
t
(37n), and 8.14eV for
s
9
43
-LiCl
i 2
+ -
(iPr2N)2P-C=N-N-P(NiPrZ)Z
(iPr2N)2P-C-Li + (iPr2N)2PC1
46n (85%)
Scheme 10. Synthesis of the first stable nitrile imine 4611.
-
CH,-CEN-NSiMe,
4. Synthesis of Stable Nitrile Imines
4.1. Starting from Lithium Salts of Diazo Compounds
As often happens, the discovery of the synthetic route later
used to prepare most of the stable nitrile imines was unexpected.
In one attempt to synthesize C-acylated phosphinodiazomethane 42 d by reacting pivaloyl chloride with the lithium
salt 40 of bis(diisopropylamino)phosphinodiazomethane, the
formation of 1.3.4-oxadiazole 41 was observed in addition to
the expected diazoketone 42 d.[451Moreover, the reaction of
pivaloyl chloride and the thiophosphinoyldiazo analogue 43
quantitatively afforded the five-membered ring heterocycle
44.r461
These results strongly suggested that N-acylation could
compete with C-acylation in the case of phosphorus-substituted
diazo lithium salts, since the simplest explanation for the formation of oxadiazoles 41 and 44 was a l ,5-electro~yclization[~~~
of
the initially formed carbonylnitrile imine 45n (Scheme 9).
(see Fig. 2)), is not air sensitive, and since it melts at 100 C
without decomposition, the compound is quite stable thermally.
The influence of both the lithium salt and the electrophile on
the reaction has been studied. The reactions of the thiophosphinoyl diazo lithium salt 43 with chlorophosphanes as electrophiles were u n e q u i ~ o c a l . [ ~s~3 1. When
~ ~ - electronic factors
are constant, reactions with bulky and less hindered chlorophosphanes lead to nitrile imines and diazo compounds, respectively (Scheme 11).
s
( ~ P K Z N ) ~ P - C+ ~-N - N( N
- Pi P k - 2 ) ~
4611 (85%)
( i P r 2 N ) 2P-4 - P ( m e 2IZ
N3 4 7 d ( * )
43N2
s
tBu2PCl
s
(iPr2N)zP-$-Li
+ -
(~P~~N)ZP-C=N-N-P~BU~
48x1 ( * )
PhZPC1
S
(iP r ~ N ) ~ b - PPh2
5
N2 4 9 d ( * )
-
Ar : F3c-4F3
PMrPC1
CF3
-
S
( ~ P -T ~-N ) ~ ~ : - C ” ~ - ~ - P P ~ K
50n
R2P-C-Li
tBu-C-C1
0
4
41(25%)
S
R&C -LX
(82%)
Scheme 11. Influence of the chlorophosphanes in the reactions Hith thiophosphnoyl diazo lithium salt 43. (*): High yield according to ” P N M R spectrum; the
compound was not isolated.
426 (75%)
Rzi-t:-tBu
N-N 44
N, 43
[R,f-C&-i-!-tBu
Comparison of the results of the reaction of diazo lithium
salts 40 and 43 with various electrophiles (Scheme 12) poses a
second problem. The nitrile imines resulting from the N-attack
of the electrophiles can isomerize into the thermodynamically
fwored diazo compounds. Scheme 12 gives four examples of
45n
Scheme Y. ,V- and C-acylalions of phosphorus-substituted diazo lithiurn salts.
R = ;Pr2h. X = lone pair or S.
+ -
R2P-C=N-N-Sine3
However, although salts of diazo compounds possess two
nucleophilic centers, it was generally believed that electrophiles
only react at carbon giving the corresponding substituted diazo
compound.[481Since diazo compounds and nitrile imines both
react with dipolarophiles,[61the best way to prove the hypothesis
was to try to prepare a stable nitrile imine. A good approach to
forcing the attack of the electrophiles at the nitrogen nucleophilic center is to try to prevent the attack at carbon: the use
of sterically hindered salts of diazo compounds and electrophiles was the obvious answer. Moreover, push-pull substituents could decrease the polarity of the nitrile imine skeleton
and therefore increase its kinetic stability.
The first stable nitrile imine 46n was obtained in 8 5 % yield
from the reaction of the lithium salt 43 of bis(diisopropy1amino)thiophosphinoyldiazomethane with chlorobis(diisopropy1amino)phosphane (Scheme 10)
491 Nitrile imine 4611,
which was fully characterized (including an X-ray crystal study
.L463
52n
S ine 3
25OC
d R2P-$-Sine3
2 4h
N2 52d(60%)
00
c
4h
+ -
R 28 - c -sine
4,
53d(65%)
R 2P- CsN-N-SiPhg
X
54n(80%)
Rzh-S-Li
+R $ - % - S i p h 3
25’C
N2
40.43
55n
+ -
X:lp
6h
N2 55d(90Z)
+ -
R 2P- CSN-N-BR~
58n(90%)
Scheme 12. Reactivity of diazo lithiiim salts 40 and 43 with various electrophiles.
R = rPr,N; lp = lone pair.
531
G. Bertrand and C. Wentrup
REVIEWS
this rearrangement, which takes place, depending on the substituents, between 0 and 5 5 0C.[49. 541 Therefore, one possible
way to rationalize the formation of diazo compounds in cases in
which no nitrile imines were observed was to postulate that the
nitrile imines were always the kinetic products of the electrophilic attack on the diazo lithium salts, but that they rearranged into the thermodynamically Pdvored diazo compounds
(see Section 7). This hypothesis was reinforced by the formation
of diazo compound 51 d[45.'I in the reaction of the phosphinoyldiazo lithium salt 40 with chlorobis(diisopropy1amino)phosphane, and the formation of nitrile imine 46n in the
analogous reaction of the thiophosphinonyl analogue 43.[4h*
491
Indeed. since lithium salts 40 and 43 are sterically quite similar,
the opposite regioselectivity of the reactions was difficult to
explain. The results observed are apparently best attributed to
electronic factors : the presence of two electron-donating groups
destabilizes nitrile imine 51 n (R,P-CNN-PR,,
R = iPr,N)
thus favoring the rearrangement to the thermodynamically preferred isomer 51 d.
Finally, the concomitant formation of both isomers 61 n and
61 d in the reaction of the lithium salt 60 of bis(dicyclohexy1amino)phosphinoyldiazomethane and chlorobis(dicyclohexy1amino)phosphane, and the absence of isomerization of 61 n into
61 d, demonstrated that the C- and N-attacks of electrophiles on
salts of phosphorus-substituted diazo compounds are competitive processes (Scheme 13) .[561
'.
-
+ R2PCl
+ R~P-CEN-N-PR~
N2 60
+
C-attack
N2
Scheme 15. The two pathways for the formation of diazo compounds
In any case, the route employing diazo lithium salts appeared
to be a very general method for synthesizing stable nitrile
imines, as two further examples demonstrate. The two C-silylated nitrile imines 6 2 1 1 ~"I ~and
~ . 63n["] were prepared, and more
recently the two C-borylated nitrile imines 64n and 65n
were synthesized from the first stable boryldiazomethane
(Scheme 16).[". '91
,-'
+ -
iPrgSiCl
i P r 3Si-C=N-N- S i i P r 3
62n (80%)
-2VzBc1
,
(iPr2N)2PC1
I
( i P r 2 N ) ~ B - $- L i
+ -
iPr3Si-CrN-N-B(NiPr2)2
63n (80%)
+
+ -
( i P r 2N)2B-C=N-N-P( N i Pr 2)2
64n(90%)
+ -
(iPr2N)+CsN-N-B(NiPr2)2
65n (85%)
Phosphorus-carbon bonds are not easily cleaved, thereby
limiting application of the C-phosphorus-substituted nitrile
imines in organic synthesis. In contrast, silicon-carbon and
boron -carbon bonds are easily cleaved. Thus, nitrile imines
62n-65n are not only laboratory curiosities but potential building blocks for the facile synthesis of a variety of heterocycles.
R:iPrzN,X:S
, (iPr2N)2$-CrN-N-P
s + -
(NiPr2)~
46n
(iPr2N),p-$-P(NiPr2)z
N2 51d
R: ( C-C6H11) zN; [(c-C6H11)ZN1 ZP-$-P[N
X:lp
+
N2
[(c-C6H11)
2
616
+ -
[(C-C6H,,),NIzP-C~N-N-P [N(C-C6H11)212
616
Scheme 14. Influence of the diazo lithium salt on the formation of the nitrile imine
and diazo derivative.
Although electronic factors are not totally negligible, steric
factors are of primary importance for the kinetic stabilization of
nitrile imines that can isomerize into the thermodynamically
preferred diazo compounds. Thus, the formation of diazo compounds can result either from C-attack of the electrophile o r
532
I
4.2. Starting from Stannyldiazo Compounds
I
I
4
N2 61d(iO%)
A very small change in the steric factors, such as replacing a
lone pair by a sulfur atom, or exchanging diisopropylamino by
dicyclohexylamino groups, can totally reverse the ratio of nitrile
imine to diazo compound (Scheme 14).
R2P-$-Li
E+
Scheme 16. Synthesis of c'-silyl- and C-borylnitrile imines
Scheme 13. Concomitant formation of diazo derivative 61d and nitrile imine 61 n.
x
+
N2
RzP-$-PR2
61n(90%)
,
X-f-Li
N2
+ -
R2P-f-Li
from N-attack followed by rearrangement of the initially
formed nitrile imine (Scheme 5 ) .
Since the regioselectivity of the attack of electrophiles on the
lithium salts of diazo compounds depends almost entirely on
steric factors. it was logical to think that the use of diazo precursors bearing a metal atom larger than lithium would favor the
formation of nitrile imines over diazo compounds; stannyl diazo compounds were potentially good candidates. In contrast to
lithium diazo compounds, their structures are well established;[601the tin atom is bound to carbon. The reaction of
stannyl(thiophosphinoy1)diazomethane 66d with chlorobis(diisopropy1amino)borane was examined first. Indeed, the corresponding nitrile imine 59 n, already prepared by the diazolithium salt method, was obtained.[561However, no reaction occurs
between stannyldiazomethane 66d and chlorobis(diisopropy1amino)phosphane, thus proving that 66d is less reactive than its
lithium analogue 43 (Scheme 17).['61
In the same way. triisopropylsilyl(trimethylstannyl)diazomethane (67d) does not react with triisopropylchlorosilane,
but a clean reaction occurs with chlorobis(diisopropy1amino).4nges Clieni In1 Ed Engl 1994, 33, 527-545
REVIEWS
Nitrile Imines
(90%)
(85%)
+ -
R3 S i-CaN-N-5 i R 3
MegSn-5-SnMe3 R3SiC1
___) ~e3Sn-C.~-n-SiR3]R3sic1
46n
N2 696
71n
RgSi-$-SnMe3
R : iPrZN
R2BCl
R :i P r
R2BCl
59n
in solution
(55%)
(95%)
(60%)
N2
676
Scheme 17 foinparison of the reactivity of stannyl- and lithiodiazo compounds
66d and 43.
Scheme 20. Reactivity of 69d with chlorotriisopropylsilane.
phosphane leading to nitrile imine 68n.[s61 This result is of special interest since nitrile imine 68n was not available by the
diazolithium salt method, which instead gave the diazo isomer
68d[54' (Scheme 18).
is certainly N-silylation giving rise to the N-silyl-C-stannylnitrile
imine 71n. In the absence of solvent, 71n undergoes a substitution reaction by another molecule of chlorosilane affording bis(sily1)nitrile imine 6211, whereas in the presence of a solvent, the
nitrile imine isomerizes to the diazo compound 67d
(Scheme 20).
These results seemed to show that, in contrast to lithiodiazomethanes, stannyldiazomethanes always react with electrophiles at the nitrogen terminus. However, it has been proven
that they also undergo substitution at the carbon atom. Diazo
compound 69d reacts with the bulky pivaloyl chloride giving
oxadiazole 73, thus implying the primary formation of N-acylC-stannylnitrile imine 7211;in contrast, with the sterically less
demanding acetyl chloride, acyl(stanny1)diazomethane 74d is
formed (Scheme 21). [s61
i Pr S i c1
no reaction
iPrgSi-$-SnMe3
NZ 676
+ LiPr3Si-C=N-N-PR2
R2PCl
R : iPrZN
iPr3Si-C-PR2
686 N2
R2PCl
Scheme 18. Comparison of the reactivity of stannyldiaao derivative 67d with that
of its lithio analogue.
The question then arises whether bis(trimethylstanny1)diazomethane (69d)[611would act as a CNN2- transfer agent
and thus allow the one-step synthesis of nitrile imines, especially
those not accessible by the diazo lithium salt method. The results summarized in Scheme 19 clearly demonstrate that this is
possible. The formation of the bis(bory1)nitrile imine 64n in
high yield shows the efficiency of the method."61 By the diazo
lithium salt method, diphosphinonitrile imine 51 n was not
available.[451while 61 n was obtained as a mixture with its diazo
isomer 61 d.[s61The stannyldiazomethane starting material unequivocally leads to nitrile imines 51 n and 61 n. demonstrating
the high selectivity of the method.[s61The most striking result is
certainly the preparation of the first stable, purely organic nitrile
imine 70n[56.621(Scheme 19).
0
,\
Be 3 S n -C- SnMe
73 (57%)
696 N2
N2 74d(90%)
0
Scheme 21. Reactivity of 69d with acyl chlorides
Stannyldiazo compounds 67 d and 69 d have also been treated
with germanediyl75 to prepare germyl-substituted nitrile imines
76n and 77n.[631However, these reactions proceed by a completely different mechanism, since no trimethylchlorostannane
is eliminated; so far this mechanism has not been investigated in
more detail (Scheme 22).
( ~ P ~ Z N ) ~ B - C+. N-- N - (BN Z P X - ~ ) ~
iP r3S i-C- SnRg
676 N2
6411(90%)
(R ~ N ) ~ P - C S & P ( N
696
1
Ph3CC1
,
R2)2
51n. 61n( 95'90%)
+ -
Cp*Ge-R'
75
Scheme 19. Synthesis of nitrile imines from bis(stanny1)diazomethane 69d.
The dramatic differences observed in the reaction of 69d with
triisopropylchlorosilane, with and without solvent, are of primary interest (Scheme 20) .[s61 Since silyl(stanny1)diazomethane
67d does not react with chlorotriisopropylsilane, it is obvious
that 67d is not the intermediate leading to nitrile imine 62n.
Thus, under both sets of experimental conditions the first step
R:Me
R':CH(SiMe3)2
b
+ -
SnR3
I
iPr,Si-CaN-N-Ge-Cp*
76n
k'
*
Ph3C -C EN-N-CPh3
51n R : i P r
A n g m Clieni In1 Ed Engl 1994. 33. 521-545
I
R3Sn-$-SnR3
696 N2
CP - q e -C aN-N- G,e-Cp
R' 77n R '
Scheme 22. Synthesis of nitrile imines starting from germanediyl 75. Cy'
C5Mes
*
=
Stannyldiazo compounds allow the straightforward synthesis
of a variety of symmetrically and unsymmetrically substituted
nitrile imines. The experimental conditions are very mild, and
the trimethylstannyl chloride formed is easily removed by sublimation
mm Hg, room temperature). Moreover.
533
G. Bertrand and C. Wentrup
REVIEWS
bis(trimethylstanny1)diazomethane (69 d) can be easily prepared
in multigram quantities and acts as a safe CNNZ- transfer
reagent, in contrast to analogues such as the dimercury, disilver,
and dilithium diazomethane salts, which are e x p l ~ s i v e . ~ ~ * ~
4.3. Starting from Stable Nitrile lmines
+ -
s
(98%)
+ -
MeOTf
R2fi-C.N-N-PR2
s
s8
+
5811
+
46n
(87%)
+ -
XBQ
R2i!-CEN-N-PPhAr
50n
+
(50%)
s
+ -
R2fi-C.N-N-BR2
5 9n
+
59n
+-
1) H e L i
2 ) EC1
s
+ -
R~~s-c.N-N-E
-R2BHe
-LiCl
E : RZP 46n (68%)
E : S i i P r j 57n (85%)
R: i P r 2 N
Scheme 24. N-substitutions on a nitrile imine.
Since stable nitrile imines were available, the question arose as
to whether it was possible to use the heteroatom bound to the
1,3-dipole as a second center of functionality in the molecule. In
other words, was it possible to d o chemistry at the periphery of
the nitrile imine without destruction of the C N N skeleton?
Three examples have been reported with phosphinonitrile
imines. Addition of elemental sulfur to the N-boryl-C-phosphinonitrile imine 58n led quantitatively to the C-thiophosphinoyl
already obtained by the lithiodiazomethane
analogue 59n,r581
and the stannyldiazomethane methods (Scheme 23). More in-
R2P -C.N-N-BRz
S
R,#-C&-BR~
-+
reaction must be electrophilic attack by the boron atom, leading
to a transient borate 81, which after elimination of bis(diisopropy1amino)methylborane gives rise to diazo lithium salt 43
(cf. Scheme 9).
Interestingly, formal C-substitution can also be performed
with C,N-disubstituted nitrile imines.[661 Indeed, addition of
methyllithium to a T H F solution of C,N-[bis(diisopropylamino)boryl]nitrile imine (65n), followed by addition of
chlorophosphane, afforded C-phosphinoylnitrile imine 58 n in
70% yield. It has been shown that 58n does not result from
substitution at the carbon end of the nitrile imine but from
rearrangement of the first-formed boryllithiodiazomethane
82 into the isomeric "boryllithioisodiazomethane" 83r5'1
(Scheme 25)
R~P-CSN-N-PRZ TfO78n He
+
db
R2k-C.N-N-PPMr
79n
9
Si02
$1
t RzP-C=N-N-PPhAr
-liC1
80
+ -
RzB -CEN-N-BRz
65n
o"0
I
I)
HeLi
2 ) R2PCl
+ -
______)
RZP-CEN-N-BR~
-RZBHe
-LiCl
58n (70%)
Scheme 23. Synthesis of stable nitrile imines starting from nitrile imines.
Scheme 25. Formal C-substitutions on a nitrile imine
terestingly, addition of methyl trifluoromethanesulfonate to the
N-phosphinoylnitrile imine 46n afforded the corresponding N phosphonio-substituted nitrilium betaine 78n.[641 Since it is
rather difficult to imagine another
to compound 7sn, this
approach is of real synthetic interest, Lastly, N-phosphinonitrile
imine 5on, which has a phenyl and a 2,4,6-trifluoromethylphenyl group at phosphorus, reacted with tetrachloro-o-benzoquinone (TCBQ) to afford the only known ).'05-phosphorussubstituted nitrile in,ine 79,,,[651 H
~ this reaction
~
is ~not
general (see Section 7).
Surprisingly, 79n can also be obtained by simply filtering a
solution of the corresponding hydrazonoyl chloride 80 on silica
gel. The ready elimination of HCI has been explained by the
structure of compound SO and the fact that the proton and the
chlorine atom are only 2.45 8, apart (according to a singre-crystal X-ray analysis).[651So far, 79n is the only example of a stable
nitrile imine prepared by this classical method.
The high reactivity of boron-nitrogen bonds has been exploited in substitution reactions of boryl-substituted nitrile
imines.[6"1 Addition of methyllithium to N-boryl-C-thiophosphononitrile imine 59 n, followed by addition of chlorobis(diisopropy1amino)phosphane (or chlorotriisopropylsilane) led to the
corresponding N-phosphinoylnitrile imine 46n (or N-silylnitrile
imine 57n) in reasonable yields (Scheme 24). The first step of the
534
By using these properties, it has been possible to substitute
nitrile imines at both carbon and nitrogen in a two-step sequence. C,N-[Bis(diisopropy'aminofboryllnitrile imine (65n)
was first treated with methyllithium and then with chlorobis(dicyclohexylamino)borane. This led to C-[bis(dicyclohexylamino)boryl]-N-[bis(diisopropylamino)boryl]nitrileimine (84 n) ,
The
sequence
was
~ reaction
~
~
, repeated with 84n to give C,N[bis(dicyclohexylamino)boryl]nitrile imine (8511) in fair yield
(Scheme 26) .[661 This method could be of special interest when
1)lfeLi
2)R '2BC1
+ R2B-CsN-N-BR2
6511
R:iPr,N
R'
__*
(72%)
+ R'2B-CSN-N-BR2
84n
1)HeLi
2)R'zBCl
(85%)
+ R'ZB-C'N-N-BRi
85n
: (C-C6H11)2N
Scheme 26. Formal C- and N-substitutions on a nitrile imine.
the diazo compounds are not stable, since preparation of the
corresponding lithium salts would not be required. In summary,
a large variety of stable nitrile imines is now available (Table 1 ) .
Their stability is essentially due to steric factors.
Angew. Chem. Int. En'. EngI. 1994, 33, 527-545
REVIEWS
Nitrile Imines
Table I . Selected data of some nitrile imines L.X-C-N-N-YL,
1
(rPr,N),P(S)
P(NiPr,),
46x1
m.p. 1OO'C
2040
2
(rPr,N),P(S)
PIBu,
48n
solution [d]
2060
3
(iPr,N),P(S)
PPhAr [el
50n
oil
2138
4
(iPr,N),P(S)
P(NiPr,),TCBQ [el 79n
m.p. 166'C
2162
61.04
(99.4)
- 173.6
(107)
65.53
(133.7)
69.83
-215.0
(90.3. 17.7) (200)
+35.4
(5.2)
+31.6
(<I)
f27.9
(<I)
27.8
(8.6)
+28.7
(6.7)
+32.5
+31.1
+ 33.1
+
5
(iPr,N),P(S)
P(NiPr,),Me+
7811
m.p. 1OO'C
2170
6
7
8
(iPr,N),P(S)
(iPr,N),P(S)
(iPr,N),P(S)
SiMe,
SiPh,
SiiPr,
5311
55n
5711
solution [d]
solution [d]
m.p. 54°C
2010
2120
2050
9
(iPr,N),P(S)
B(NiPr,),
5911
oil
2095
55.67
(132.2)
f31.8
10
(iPr,NI2P(S)
BMes, [el
m.p. 84 C
2144
+27.1
11
(iPr,N),P
P(NiPr,),
51n
oil
2047
12
[(IPC,H,,),N)],P
P[N(c-CeH,,)J2
6111
oil
2047
71.48
(131.2)
63.28
(62.5)
64.40
(72.8)
13
14
15
P(NiPr, j2
P(NiPrl),
P(NiPr,)2
SiMe,
SiPh,
SiiPr,
5211
solution [d]
oil
oil
2100
2140
2110
b.p. 11 0 "C
(0.2 Torr)
solution [d]
b.p. 90 ' C
(0.05 Torr)
m.p. < 20 "C
oil
m.p. 136'C
b.p. 105°C
(0.001 Torr)
b.p. 110°C
(0.001 Torr)
m.p. 115'C [t]
2113
2152
5411
5611
16
17
18
19
20
21
22
23
24
25
[(c-C,H,,),Nj1B]
B[N(c-C,H,,),],
85n
m.p. 121 "C [fl
26
Ph,C
CPh,
70n
m.p. 60°C
[q
61.87
(48.3)
2099
2120
46.73
2164
2076
2052
2145
45.10
49.83
63.71
67.00
2160
2155
2052
-191.9
(250)
65.80
68.01
(175)
66.50
(550)
85.5
(280)
-189.0
(78)
- 182.9
- 173.0
- 184.4
(220)
- 186.4
(140)
- 188.0
+45.5
(9.0)
52.0
(9.1 )
+44.1
f42.4
f45.7
+
85
68
+119.0
A
C
A
+70.6
A
82
-24.3
C
C
50
90
87
+ 27
A
A
A
C
A
B
C
A
96
85
95
60
98
65
f96.0
B
95
+100.0
B
90
- 10.4
A
A
A
80
90
f99.9
C
f52.7
+12.8
+45.1
f6.6
(3.6)
29
+0.71
f101.4
f6.4
+8.35
+ 22
+
+ 22
+96.6
+ 23
+ 28
A
C
B
A
B
A
B
B
A
60
48
90
90
71
90
70
80
90
xo
X5
+ 27
+ 27
A
B
C
-191.0
+25
+ 25
C
85
-182.1
60.71
16.26
B
90
[a] External standards "C and 29Si N M R : Me,%; "B. I4N, and ,'P N M R : BF, . OEt,. MeNO,, and H,PO,, respectively. Jpc and Jpxcoupling constants in Hz; L~~~~ in
Hz. [b] A, B. C refer to the synthetic methods described in Sections 4.1.4.2, and 4.3, respectively. [c] Isolated yield. [d] Compounds characterized in solution but not isolated.
[el Ar: 2.4.6-tris(trifluoromethyl)phenyl; TCBQ: tetrachloro-o-benzoquinone; Mes: 2,4,6-trimethylphenyl; R,R',R': pentamethylcyclopentadienyl, trimethylstannyl.
bis(trimethylsi1yl)methyl. [fl Decomposition
5. Characterization of Nitrile Imines
5.1. X-ray Data and Ab Initio Calculations
The key angles and bond lengths in those nitrile imines for
which X-ray crystallographic structures have been determined
are listed in Table 2. Selected X-ray structures are shown in
Figures 1-3. Inspection of the data immediately reveals the
large variability of structures, which range from close to propargylic (cf. 2 a) to allenic (cf. 2 b) .
The X-ray data in Table 2 may be compared with the results
of high-level a b initio calculations[121on model compounds
(Table 3). On the basis of STO-3G and HF/4-31G calculations
Houk et
' I 1 found two minima for the unsubstituted nitrile
imine H C N N H 2n a t the STO-3G level. These correspond to 2a
and 2b. However, at the MP2/6-31G* level, there is only one
stable form, which is bent and nonplanar and thus resembles the
allenic structure 2b more than the propargylic 2a. This preferA i q y C h m i Inr Ed EngI. 1994. 33, 527-545
ence for the allenic form is further supported by higher level
geometry optimization at the MP2/6-311 + G(2df, 2p) and
QCISD/6-311 G** levels (Table 4) .['*I
t
-
Fable 2. Selected X-ray data for nitrile imines X-C-N-N-Y
[a]
Parameter [a]
70n
46 n
78 n
85 n
77 n
r(C-N)
r(N-N)
r(X-C)
r(N-Y)
<(XCN)
<(CNN)
<(NNY)
I(XCNN)
r(CNNY)
lbl
ref.
1.173
1.262
1.503
1.530
137.3
169.4
115.2
135.7
134
2052
1621
1.177
1.240
1.771
1.777
138.2
173.6
115.0
121
152
2040
[491
1.143
1.285
1.815
1.623
163.5
169.6
123.6
84.1
177.7
2170
[64 bl
1.171
1.268
1.559
1.452
165.1
169.4
122.4
52 48
-178 9
2160
[661
1.183
1.264
1.935
1833
150.1
174.2
129.9
68.2
148.7
2052
[631
[a] Angles in degrees, bond lengths in A. [b] Frequencies [cm-'1 of the IR absorption band of the C N N unit for comparison.
535
G. Bertrand and C. WentruD
REVIRNS
Table 3. Calculated structural parameters (MP2.'6-31G*) for C- and N-substituted
nitrile imines XCNNH and HCNNY [12].
Suhstiturnr
Pard-
X
meter [a]
H
PH,
PH:
BH,
SiH,
CH,
C,H,
r(C-N)
r(N-N)
v(X-C)
r(N-H)
3:(CNN)
Qr(XCN)
h(HNN)
r(XCNN)
r(HNNC)
I(CNj {bl
1.192
1.267
1.073
1.027
169.2
140.6
108.0
138.1
131.8
2383
1.194
1.268
1.777
1.028
169.6
148.6
108.0
134.4
150.3
2442
1.199
1.227
1.705
1.029
170.0
149.0
116.2
112.4
348.2
2399
1.190
1263
1.485
1.030
172.3
178.7
107.8
179.9
179.7
2485
1.194
1.263
1.825
1.026
171.5
175.1
309.2
108.6
162.1
2431
1.193
1.273
3.476
1.028
169.1
145.4
107.6
134.3
135.9
2460
1.195
1.270
1.436
1.028
169.4
150.3
10X.l
127.5
137.0
[cl
Parameter [a]
H
PH,
PH:
BH,
SiH,
CH,
C,H,
r(C-N)
r(N-N)
r(H-C)
r(N-Y)
Qr(CNNj
3:(HCN)
X(YNN)
T(HCNN)
T(YNNC)
i(CN) [b]
1.192
1.267
1.073
1.027
169.2
140.6
108.0
138.1
131.8
2383
1.196
1.258
1.072
1.757
170.9
144.8
115.6
151.2
114.0
2319
1.172
1.289
1.074
1.641
170.8
179.9
119.4
171.0
179.9
2217
3.197
1.257
1.068
1.422
166.3
151.5
121.8
180.0
180.0
2146
1197
1.250
1.070
1.755
171.1
147.0
120.4
145.2
120 1
2315
1.196
1.260
1.075
1.474
169.8
138.1
111.8
153.2
114.6
2351
1.195
1.262
1.076
1.419
169.7
140 1
115.2
148.8
Subuitucnt
A
Y
Fig. 1. Crystal structure of 7011[62]
118.1
[c]
[a] Angles in degrees. bond lengths in A. [b] C N N stretching frequency in cn7C'
(uncorrected MP2i'6-31G* values). Note: these vibrations are coupled with the
N - H or C - H stretching modes and therefore cannot be compared with experimental values for differently substituted compounds. [c] Beyond the computational
power of current supercomputers.
Table 4. Calculated structural parameters of formonitrile imine 2n (allenic form
- +
H C = N = N - H 2a) [12].
Parameter
HF,'6-31G* MP2/6-31G* MP2;6-311
r(C-H)
r(C-N)
r(N-N)
r(N-H)
3:(HCNj
3: (CNN)
3::(HNN)
r(HCNN)
r(HNNC)
[a] QClSD
tions.
1.069
1.179
1223
1.009
127.5
171.7
108.7
140 2
130.1
=
1.073
1.192
1.267
1.027
140.6
169.2
108.3
138.1
131.8
Level
+ G (2df,?p) QCISD,631 lG** [a]
1.065
1.181
1.253
1.018
145.5
170.6
108.8
132.6
137.2
1.082
1.207
1.254
1.028
128.2
168.5
107.1
140 8
130.6
Quadratic configuration interaction with single and double excita-
The MP2/6-311G**(2 df. 2p)-level geometry of HCNNH (Tables 3 and 4) is in excellent agreement with the experimental
data for bis(triphenylmethyl)nitrile imine (70n) given in Table 2
and Figure 1 . Thus, this compound may be considered typically
allenic and perturbed minimally by steric or electronic effects.
The N-phosphino-C-thiophosphinoylnitrile imine 46n
(Fig. 2) also has an allenic structure very similar to that of 7011;
however, the structure of the phosphonio-substituted compound 7811is dramatically different (Fig. 3). The P-C-N angle is
much flatter (163.5'); the C-N bond is particularly short
( 1 .I4 A), and the N-N bond correspondingly long (1.285 A).
This change in geometry is nicely corroborated by the results of
calculations for HCNNPH: (Table 3). which was shown to be a
planar and "propargylic" compound with a linear HCN moiety,
a short C-N bond, and a long N-N bond. The main difference
536
Fig. 2. Crystal structure of 46n [49]
Fig. 3 Structure of the cation of 7811in the crystal [64b]
between the actual structure of 7811and that calculated for the
model compound is the dihedral angle of 84" between the PCN
and CNN planes in the former. However. the CNNP moiety in
7811is nearly planar, and the CNN backbone nearly linear, so
that the deviation from the calculated structure is not so dramatic, and it may well be caused by steric and/or crystal lattice
effects.
According to calculations the boryl compound
H,B-CNN-H
also tends toward linearity and planarity, although the C-N bond is not quite so short and the compound
thus less "propargylic" (Table 3). The bis(bory1) compound
H,B-CNN-BH,
has a similar structure."'] These trends are
A n g e ~ Chem.
.
Int. Ed. Engl. 1994. 33, 527-545
REVIEWS
Nitrile Imines
in accord with the X-ray structure of 85n, which is similar to
that of the phosphonio compound 78n but not quite as clearly
propargylic as judged by the lengths of the C-N and N - N
bonds (Table 2). The BCNN skeleton is close to linear
(BCN = 165.1 ’), and the B-C bond is quite short (1.559
In the propargylic forms of these molecules with partial C = N
triple bonds the negative charge on the terminal nitrogen atom
is increased as in 2a, and the electronegative N-phosphonio and
N-boryl substituents are thus stabilized. Accordingly, the N - P
bond in 78n is rather short (1.623A in 7811 vs 1.777A in
46n)
The calculations indicate, however, that a BH, group
on carbon has a greater “linearizing” effect than one on nitrogen (Table 3) ,[’
A).
’’
entry 9), but the more electrophilic dimesitylboryl substituent in
entry 10 has a large effect.
None of the stable nitrile imines for which X-ray structures
have been determined so far are truly “propargylic” : none have
a linear X-C=N- backbone, and none absorb in the actual
nitrile region of the IR spectrum. It appears that the C-arylnitrile imines examined in low-temperature matrices constitute a
separate class with nitrile-like Ar-CEN- moieties, which absorb above 2200 cin- in the TR spectrum.
Diphenylnitrile imine (11 n) absorbs at 2228 cm-’ in PVC
matrix[371and at 2242 in Ar matrix.[691 15N labeling of the
central nitrogen atom causes, as expected, a large isotopic shift
(to 2195 cm-I), whereas labeling of the terminal nitrogen atom
has a miniscule effect (2227 cm-’ in PVC matrix).1371This is in
f
5.2. Infrared Spectra
The frequencies of the asymmetric C N N stretching vibrations
of all the stable nitrile imines are given in Table 1, and the
calculated IR spectrum of H C N N H is in Table 5. By applying
the usual correction factor[681of 0.93 to the calculated frequency, it is seen that the absorption band of the still unknown[’71
parent formonitrile imine may be expected in the range 21002200 cm-‘.
Tdble 5. Calculated harmonic vibrational frequencies and IR intensities for formonitrile imine 2 n (MP2:6-311G (2df, 2p)) [12].
No.
i-[cm-’]
503
527
620
696
115x
1253
--_
2728
3736
3296
Int. [ k m r n o l ~ ’ ] Assignment
6
36
191
46
133
28
194
95
34
C N N bend
C N N bend
C H bend
CH + N H rock
C N N antisymmetric stretch
N H bend
C N N symmetric stretch
CH stretch
NH stretch
’,
N
The experimental data in Table 1 indicates that stabilized nitrile imines fall in two classes: l ) those absorbing in the range
2 0 1 0 - 2 1 0 0 ~ m - ~ , and 2) those absorbing from 2140 to
21 70 cm - I . According to a valence bond description, the
propargylic structure - C r N - N - is expected to have a high
C = N stretching frequency approaching the nitrile region
( 2 2 2 0 0 cm-I), whereas the frequency for the allenic structure
- C = N = N - should be lower. The data agrees with the results of
the X-ray determinations described previously. Thus, the compounds listed as entries 3 - 5 11, 14. 19, and 22-25 in Table 1
may be described as partially propargylic. For compounds 78 n
and 85n (entries 5 and 25) this has been discussed already. The
electron-withdrawing tris(trifluoromethy1)phenyl substituent
on phosphorus in 50n (entry 3 ) is seen to have a large effect and
increases the CNN stretching frequency into the “propargylic”
range. Even phenyl groups on silicon have this effect in compounds 5511and 54n (entries 7 and 14). Boron has a large effect
in the same direction both in C-boryl (6411, entry 22) and C,Nbis(bory1) compounds (6511, 8411, 8511, entries 23-25). The diaminoboryl substituent on nitrogen alone has little effect (5911,
-
line with the nitrile-type Ar-CZN-N-Ar
structure. The corresponding 15N isotope shift in benzonitrile itself is 27 cm-‘
( P h C e N 2228 c m - l ; P h C r ” N 2201 cm-’ in PVC).’371 Evidence that the compound observed was in fact diphenylnitrile
imine was given by its photochemical rearrangement to
diphenylcarbodiimide and cleavage to benzonitrile and phenylnitrene (see Section 7).[371
Similarly. C-phenyl-N-trimethylsilylnitrileimine absorbs at
2230 cm-’ (film, 77 K), and its structure was confirmed by
mass spectrometry, photoelectron spectroscopy, and trapping
reactions,143.441as well as by its photochemical rearrangement
to N-phenyl-N’-trimethyIsilylcarbodiimide.[6y1
Dimesitylnitrile imine (25n) (Scheme 5) absorbs in solution at
2200 cm- as Grundmann and Flory observed.[’”]. However,
N-methyl-C-phenylnitrile imine, generated by Ar matrix photolysis of 2-methyl-5-phenyltetrazole at 12 K , absorbs at 2032
and 1361 cm-’ and is photochemically converted to N-phenylN’-methyl-~arbodiimide.~~~]
It would be highly interesting to
synthesize a stable C,N-diarylnitrile imine to further investigate
the structure.
N-Phenylnitrile imine (87n), generated in PVC matrix by
photolysis of 2-phenyltetrazole (86) [Eq. (91, was reported to
hv
H-{*y
___)
N-N,
86
+ -
H-CEN-N-Ph
87n
(5)
Ph
absorb at the low frequency of 2014 cm-1.[371This compound
is a derivative of HCN, which absorbs near 2100 cm-’.[701Nitrile imine 87n, which lacks a C-aryl substituent, may be more
of the allenic type. 15N labeling could give more information,
but at any rate a low-frequency absorption band for 87n is not
+ unexpected. The nitrile oxides R-CEN-0, which are linear or
nearly linear,”
absorb at very high frequencies near
2300 cm-’, except for the parent member, fulminic acid.
+ H-C=N-O, which has a strong band at 2198 cm-’ (gas)[’81o r
2193 cin- (Ar matrix) .[721
It is instructive to note that the C z N - C absorption frequencies of nitrile ylides span almost 400 cm-’. Benzonitrile
methylide 89, generated in Ar matrix by photolysis of 2 H azirine 88, was reported to absorb strongly at a relatively low
frequency, 1930 ~ m - ’ . [ ’ ~The
]
only stable nitrile ylide known
was prepared and characterized by Arduengo et al.:[741in com537
G. Bertrand and C. Wentrup
REVIEWS
pound 90 an essentially linear C - N z C - moiety was established
by X-ray crystallography. The IR spectrum of this compound
features a C - N stretching vibration at extremely high frequencies: a doublet at 2210, 2323 cm-1.[753Electron-withdrawing
substituents on the ylidic carbon atom are expected‘”’ to stabilize the linear structure as in 90 and 91. Ylide 91 has been isolated in matrix and absorbs in the nitrile region at 2250 cm- [’‘I
(Scheme 27).
Scheme 28. Photolysis of 2H-azirine 92.
that diphenylnitrile imine is linear because of the bathochromic
shift of the UV absorption maximum relative to that of benzonit
88
89
+
-
= 344 nm). If 11 n were
trile benzylide (PhCEN-CHPh; inax
bent like 94. a hypsochromic shift would be expected. Nevertheless, the high-level ab initio calculations given in Table 3 predict
that nitrile imines with aromatic substituents will be nonlinear
and allenic.
- CF3
~BU-CSN-C<
F3C
CF3
5.4. NMR Spectra
CF3
91
90
Scheme 27. Nitrile ylides characterized by I R spectroscopy. Ad
=
adamantyl
’
The span of approximately 200 cm- in the symmetric
stretching vibration vCNNin nitrile imines is thus perhaps not so
surprising. Unfortunately, the calculated IR frequencies
(Table 3) cannot be compared directly with experimental values
for differently substituted compounds, because the vCNN
vibrations in the monosubstituted nitrile imines are found computationally to be strongly coupled with either the C-H or N - H
stretching modes. The existence of coupled vibrations could
explain all the variations observed in the IR spectra. More work
is needed to understand the relationship between structure and
vibrational frequencies in nitrile imines.
5.3. UV Spectra
Selected UV spectral data for nitrile imines are given in
Table 6. While less structural information can be gleaned from
these data, it is instructive to compare the UV spectrum of
diphenylnitrile imine (1 1n; L,,, = 378 nm) with that of the cor-
Table 6. UV spectroscopic data of nitrile imines.
Cmpd.
L.,,
[nml
Phase
Ref.
26 n
33 n
378
383
256. 292
244, 273
275
272
310
EPA or PVC
EPA
pentane
pentane
pentane
pentane
PVA
(36%371
(281
~541
54 n
56 n
57 n
62 n
87 n
[541
(541
P I
[371
+ responding benzonitrile benzylide PhC=N-CHPh.L771In their
discussion whether nitrile ylide 93 or 94 is generated by photolysis of 2H-azirines 92 in 2 M P or DMBP glasses at 77 K
(Scheme 28), Hansen and Heimgartner’”] argue in favor of the
bent allene-type structure 94, because the changes in the UV
spectra are more in line with those observed for benzylideneamines than for benzonitriles. In the same vein. one can argue
538
In contrast to X-ray analyses and IR spectroscopy, N M R
spectroscopy does not really allow a differentiation between the
allenic and propargylic structures. For illustration, we can compare the N M R data for the best representative of each structure
type, namely the allenic N-phosphino-C-thiophosphinoylnitrile
imine 46n and the propargylic N-phosphonio-C-thiophosphinoylnitrile imine 78n. The I3C N M R signal for the quaternary
carbon atom of 46n and 78n appears at 6 = 61.04 and 69.83.
respectively; the difference is not significant. The I4N NMR
signals of46n and 76n, 6 = - 173.6 and -215, respectively, are
possibly more informative. It is, however, surprising to find a 6
value of only - 191.0 for the C,N-bis(bory1)nitrile imine 8511.
However, N M R spectroscopy is a very valuable tool for the
identification of nitrile imines. In studies of reactions in which
a nitrile imine andjor a diazo compound can be formed, a rapid
and reliable method is needed for distinguishing these two structural isomers. The IR absorptions for both types of compounds
are very similar.
The 13C N M R chemical shift of the quaternary carbon atom
of nitrile imines is generally in the range 6 = 45-70; ditritylnitrile imine with a signal for the corresponding carbon atom at
6 = 85.5 is an exception. This chemical shift range can serve as
a criterion for differentiation of nitrile imines and diazo compounds, since the signal of the diazo carbon atom appears at
higher field (6 < 40). However. observation of these 13CN M R
resonances is hampered owing to the fairly long longitudinal
relaxation time TI ( I 3C).In addition, the transverse relaxation
time Tz (‘3C) is considerably shortened by second-order scalar
relaxation caused by the adjacent I4N nucleus. Further difficulties arise when nuclei such as “ B or 31P are attached to the
carbon atom, giving rise to additional broadening or splitting,
respectively.
So far, all attempts to obtain a I5N N M R spectrum of nitrile
imines have failed. In contrast, 14N N M R spectroscopy provides a simple and quick way to identify nitrile imines and differentiate them from diazo isomers.‘781The I4Nz nuclei (CN,N,)
give rise to comparatively narrow 14N N M R signals, which are
easily observed at room temperature. The chemical shifts of the
N, atoms of nitrile imines are between 6 = - 170 and - 21 5 ,
while for diazo compounds the analogous signals are between
6 = - 1 10 and - 130; this difference is sufficient for a definitive
structure assignment. Note that, as expected, the I4N2chemical
A n g m Chein. Int. Ed. EngI. 1994. 33. 527-545
Nitrile Imines
REVIEWS
Lastly, it should be noted that the reactivity of the N-borylnitrile
imines (58n,59n, 63n) is quite usual but. surprisingly, the C'-borylnitrile imines (6411,6511.84n, 85n) are rather ~nreactive.['~]
1,3-Dipolar cycloaddition reactions are usually stereospecific,
and it was long assumed that any exception was due to prior
isomerization of the dipolarophile or subsequent isomerization
of the product.[4% Surprisingly. nitrile imine 46n reacted
stereospecifically with dimethyl fumarate affording the tram
adduct 95, while with dimethyl maleate a mixture of civ-96 and
t r i m adducts 95 was obtained. Since attempted epimerization of
96 failed, and N M R spectroscopy revealed that the unconsumed
dipolarophile was not isomerized when the reaction was run
with 1.2 equivalent of maleic ester, it was concluded that the
lack of stereoselectivity observed was due to a "nonconcerted
addition" process[49] (Scheme 30). Two previous examples of
shifts for the nitrile imines are found in a range similar to those
of nitrilc oxides.[791The 14N, resonance has never been observed even at high temperature (80 ' C ) , probably because the
electric field gradient at the site of the I4Npnucleus is very large,
thus causing the l4N# relaxation time T, to be very short.
Since many of the known, stable nitrile imines have heteroatom substitutents, N M R data from heteronuclei have been
widely used for structure determination. In the case of C-phosphorus-substituted nitrile imines, the 3 ' P chemical shift was
always at higher field than that of the diazo isomer; this is
particularly dramatic with thiophosphinoyl substituents
(Ah = 27-43) but also evident with phosphino substituents
(Ad = 9 12). Moreover, the presence of a phosphorus atom at
the carbon terminus induces couplings with the nitrile imine
carbon ('Jpc-= 48-133 Hz) and also sometimes with the N-nitrile imine substituent (4Jpp= 0-9 Hz; 'JPsi= 0-4 Hz). The
'9Si N M R signals are somewhat difficult to observe and are not
indicative of the structure. In the same way. "B N M R spectra
are not very helpful, since the signals are very broad and the
range very narrow. For instance, in the case of C,N-bis(boryl)nitrilc imine 6511,the boron atoms can only be differentiated
at 348 K ( S = f 2 3 and +28).[581
~~
95
s
+ -
R.~#-C=N-N-PR~
46n
+
95 (SOX)
R:iPrZN
R ' : C02Ke
6. Reactivity of Nitrile Imines
Scheme 30. An example of nonstereoselectivity in [ 2
6.1. 1,3-Dipolar Cycloaddition Reactions
+ 31 cycloaddition reactions
+
nonconcerted [2 31 cycloaddition reactions have been reported;
the
first was described by Huisgen et al. for the reaction of strongThe general features of 1,3-dipolar cycloaddition reactions
ly
nucleophilic
thiocarbonyl ylides with particularly electrophilic
3 R c . 771 In this section we briefly compare
are well
alkenes,[821
and
the second by Quast et al. for the reaction of
the results obtained with transient and stable nitrile iniines,
electrophilic azides with 5-alkylidenedihydrotetra~oles.[~~l
emphasizing the differences between these to types of comThe regiochemistry of the cycloaddition reactions has also
pounds.
been
rationalized in terms of frontier molecular orbital (FMO)
Nitrile imines are Class I1 dipoles in the classification of Sust5 . I 1 . 841 A s predicted, addition of unsymmetrically
tl1eory,[4.
mann;[*'I in other words. HOMO(dipole)/LUMO(dipolarsubstituted
olefins or alkynes usually yields the 5-substituted
ophile) and HOMO(dipolarophile)/LUMO(dipole)interactions
pyrazolines
or pyrazoles as the major product. However, in the
are about equally important. Consequently. nitrile imines can
case
of
C,N-bis(sily1)nitrile
imine 6211,the 4-substituted pyrabehave as both nucleophiles and electrophiles toward the dipozole
97
becomes
the
major
isomer;[s41this finding is also in
larophile; that is, the nitrile imines are ambiphilic. This stateagreement
with
theory,
since
the silyl groups raise the FMO
ment Fits well with the experimental results for transient nitrile
energies. Nitrile imine 4611reacts with methyl isocyanate only
imines. Most of the known stable nitrile imines d o not react with
with the C = N and not with the C=O double bond, as expected
styrene, butadiene. o r phenylacetylene but react with electronfor
a HOMO(dipo1e)-controlled cycloaddition and in contrast
poor dipolarophiles (note that even the transient diphenylnitrile
to
the
results observed with transient nitrile imines.'"6' (The
imine adds to dimethyl fumarate 177 times faster than to
isolated
product 99 results from the insertion of a second molestyrene) .I*'I The only real exception is the N-phosphonio-subcule
of
isocyanate
into the P-N bond of the primary product
stituted nitrile imine 78n. which is strongly e l e c t r o p h i l i ~ . [It
~~~l
98)[491
(Scheme
31).
To summarize, both transient and stable
reacts with electron-rich olefins such as ethyl vinyl ether at room
temperature, whereas the reaction with styrene requires 14 h at
60 "C. Reaction with the electron-poor olefin methyl acrylate
5'
7C
+ R'CzCH
C
needs 20 h at 80 'C (Scheme 29). This order of reactivity of 7811
RgSi-CrN-N-SiR3
)
R g S i - i \-H
+
R 3 S i - i \-R'
with dipolarophiles was confirmed by competition experiments.
' ' 3
62n
s
+ -
R2fi-C:N-N-PRz
R :iPr2N
OTf : CF3S03
R' : P h
R ' : C021ie
1Q h, 6OoC
20 h. 8 0 ° C
Scheme 29 Rextivity of nitrile imine 78n to dipolarophiles
he
46n
R: iPr
R ' : C02Me
HeN=C=O
-b
R: i P r 2 N
N-N
97(52X) 'SiR3
[
P"
s ,N.
RZfi-$ $=O
98 N-N,..]
+-
N-N,
SiR3
(48X)
HeN=C=O
S
P"
,N,
R2fi-C $=O
N-N,
99
F-I-PR2
(65%)
Scheme 31. ExdmpkS of the regloselectivity of [2 + 31 cycloaddition reaction, of
nitrilc imines 62n and 4611.
539
REVIEWS
C. Bertrand and C. Wentrup
nitrile imines have similar reactivities with 1 Jdipoles, although, not surprisingly, the stable nitrile imines are somewhat
less reactive.
6.2. Reactions Involving the CNN Skeleton and a
Heteroatom Substituent
R': co2Me
+
Scheme 33. Formal [2 41 cycloaddition reaction of niirile irnme 4611with dimelhyl acet).lenedicarboxylate.
As described in Section 5 , the structure of nitrile imines is not
strongly perturbed by heteroatoin substituents. However, one
addition and cycloaddition reactions of tetracyanoethylene usuof the important questions was whether the heteroatom bonded
ally proceed with high selectivity at the C-C double bond,["] in
to the C N N moiety could extend the scope of nitrile imine reacthis case the reaction involves the nitrile groupl"1 (Scheme 34).
tions.
Indeed, reactions involving both the C N N unit and a h3phosphorus atom bound to the nitrogen terminus have been
reported.[52,h4a1 Nitrile imine 46n reacts with sulfur, selenium,
phenyl azide, and pivaloyldiazomethane at room temperature to
afford 1 ,3,4.2h5-thiadiazaphosphole100a, 1,3,4,2h5-seIenadiazaphosphole 100 b, and 1,2,4.3h5-triazaphospholeslO0c and
100d. respectively, in good yields. Adducts 100 formally result
Scheme 34. Formal [2 + 41 cycloaddition reaction of nitrile imine 4611with tetracyanoethylene.
from a [4 11 cycloaddition. But actually a two-step mechanism
is most likely involved : electrophilic attack of sulfur. selenium.
phenyl azide, or diazo compound on the phosphorus lone pair,
Recently Regitz et al. reported a formal [2 41 cycloaddition
which is activated by the electron-rich nitrile imine moiety, leads
to intermediates 101, which undergo 1,5-ele~trocyclization[~~~reaction involving phosphinodiazomethane 105 (Scheme
35).[901All these reactions are of synthetic interest, since the
(Scheme 32). These reactions can be compared with the formaresulting heterocycles 100, 103. 104, and 106 are hardly availtion of 1.3,4-thiadiazoles and 1,3.4-triazoles from unstable Nable by classical routes.
t h i o ~ a r b o n y l - [ *and
~ ~ N-iminonitrile imines,[8hlrespectively.
+
+
s
+ -
i
R~~CEN-N-PRZ
46n
A
s
,tBu
7.
RzP-5 ;R2
tBu-5-PPh2
N-N
i:Sg
&- .
i: Se
i: PhN3
Scheme 32. Formal [l
1 O O a . X : S (85%)
1OOb. X : S e ( 8 4 %)
$N-c
,c=c;
R'
"PPhZ
106
R'
R': COZHe
Scheme 35. Formal [2
+ 4]cycloaddition reactionofphosphanyldiazomethane105.
1 0 0 ~ .X : N P h ( 7 4 % )
+ 41 cycloaddition reactions of niri-ile imine 46n.
The four reagents used in the above reactions are known to
react with the phosphorus lone pair and not with the nitrile
imine moiety. In contrast. a dipolarophile such as dimethyl
acetylenedicarboxylate could react both with the lone pair[*']
and with the 3.3-dipole. I n fact, the reaction of 4611with this
electron-poor alkyne affords the 1,2,3h'-diazaphosphinine 103
(Scheme 33) .[52, 64a1 Although heterocycle 103 formally results
from a [4 + 21 cycloaddition process, it is likely that the first
mechanistic step is the electrophilic attack of the alkyne on the
phosphorus lone pair of46n, followed by the 1,6 ring closure of
102. This is the first example of 1,6-electrocyclization involving
nitrile imines.
A similar reaction has been observed between nitrile imine
4611and tetracyanoethylene giving rise to 104.["2. 64n1 Although
540
___)
NZ 105
100
R :i P r 2 N
R'CKR'
An atypical dimerization of nitrile imines involving the N-heteroatom substituent is notew~rthy.~"]The reactions of the lithium salts of phosphinodiazomethane and thiophosphinoyldiazomethane, 40 and 43. respectively, with chlorodicyclohexylborane lead to bicyclic compounds 107 and 108, respectively (Scheme 36). It is quite likely that the first step of these
43.109b. 11Ob : X = S
Scheme 36. Atypical dimerimtion of nitrile imines 109
Angeir. Cheni. Inr. Ed. D i g / . 1994. 33. 527-545
REVIEWS
Nitrile Imines
reactions is the formation of nitrile imines 109a and 109 b, which
possess a formal boron-nitrogen double bond. Thus, the CNN
unit of one molecule can react as a 1,3-dipole with the boronnitrogen double bond of a second molecule giving intermediates
llOa and llOb having a carbocation center, which react with
the phosphorus or sulfur atoms to give the observed products.
Lastly, as already mentioned Section 4.3, the heteroatom substituents can be employed in the preparation of other stable
nitrile imines.
explain the formation of methylphenylcarbodiimide (1 10) on
(8 a) or
flash vacuum thermolysis of 2-methyl-5-phenyltetrazole
3-methyl-5-phenyl-I ,3,4-oxadiazol-2(3 H)-one (1 11) (Scheme
37).[421The yield of this reaction is rather low (1 -8 %) but can
be increased when the reaction is conducted photochemicalIY.[~’]In the thermal reaction, azine 112, which results from a
l,it-hydrogen shift. was also obtained in 10 YOyield.i421
7. Rearrangements of Nitrile Imines
In this review only the rearrangements of the nitrile imine
skeleton, and not those involving the substituents bonded to the
carbon or nitrogen termini, will be discussed.
Theoretical studies on the relative stabilities of the various
possible ground state isomers of CH,N, have been publ i ~ h e d . ’ ~Interestingly,
~]
1 H-diazirine is found to be of only
slightly higher energy than the parent formonitrile imine 2 n:
+ 14.2 kcalmol-I at the HF/6-31G
+8.4 kcalmol-’
at the HF/3-21G* level (with electron correlation),[92b1and
+ 1.1 kcalmol-’ at the G2(MP2)
Diazomethane is
significantly more stable, and carbodiimide is the most stable of
these isomers (cyanamide is the global minimum) (Fig. 4).
61
55
t’
H-C=N
-
+
H-C=N=N-H
N-N,
8a
04
-N2
He
P&-C=N-N=CH~
112
Scheme 37. Thermal rearrangement of transient N-methyl-C-phenylnitrileimine.
In a very elegant paperL3’]Holm et al. demonstrated the photochemical rearrangement of a nitrile imine to a carbodiimide.
Diphenylnitrile imine ( l l n ) was generated in PVC matrix at
85 K by photolysis of 2,5-diphenyltetrazole (8b) at 250 nm. 11 n
was characterized by IR and UV spectroscopy (v = 2228 cm- ;
i,,,=
,378 nm). Continued irradiation (A = 370 nm) of nitrile
imine l l n led to diphenylcarbodiiinide (113), along with benzonitrile (114) and l-aza-l,2,4,6-~ycloheptatetraene
(1 15)
(Scheme 38). The structures 11 n and 113- 1 15 were confirmed
by ”N-labeling experiments. Compound 115 originates from
rearrangement of the initially formed phenyl ~ ~ i t r e n e Thus,
.[~~]
the rearrangement of the nitrile imine to carbodiimide competes
with nitrogen-nitrogen bond cleavage.
’
hV
hv
250nm
375nm
’‘‘
[Ph-Czi-N-Ph]
Ph-N=C=N-Ph
Ph-iEN
___)
85
+
K
iin
HC
FH€Y
113
114
cn ..=
Fig. 4.
2 1
HN=C=NH
H~N-CZN
Calculated relative energies [kcalmol-’1 of H,CN,
isomers (G2(MP2) level (92dl).
I.
-coz
Ph-N-C=N-Me
110
N
Ph-<*y
N-N,
8b
I
0
P h - 5 F=O
N-N,
111
Me
A
+ & [Ph-CZN-N-He]
N
Ph-<*y
Therefore, considering the thermodynamics, it is not surprising
that nitrile imines rearrange to diazo compounds and carbodiimides. The nitrile imine + carbodiimide rearrangement was
postulated to proceed through a 1 H-diazirine intermediate,
which by ring-opening would lead to an imidoylnitrene; [421 the
Wolff-type rearrangement of imidoylnitrenes to carbodiimides
is well
931 Ab initio calculations indicate high energy
barriers for the ring closure of nitrile imine to 1 H-diazirine
and for the 1,3-hydrogen shift leading to diazomethane
(Fig. 4) . [ y Z b , dl Nevertheless, there is some experimental evidence
for such rearrangements.
7.1. Rearrangements of Nitrile Imines to Carbodiimides and
N - N Bond Cleavage
The thermal rearrangement of nitrile imines to carbodiimides
(both systems are isoelectronic to those in the nitrile oxide + isocyanate rearrangement)[931 has been postulated to
Scheme 38. Photochemical rearrangement of diphenylnitrile imine 11 n.
For nitrile inlines that are stable at room temperature, competing reactions of this type have only been observed on irradiation of C-[bis(diisopropylamino)phosphino]-N-triisopropylsilylnitrile imine (56n), which led to phosphinonitrile 116 and
carbodiimide 117 (Scheme 39) .[541
+ -
(ZP r 2N)2P-CzN-N-SiiPr3
56n
hv
d
( Z P K ~ N J ~ P - C Z 116
N
(10%)
+
(iPrzN)zP-N=C=N-5iZPr3
117 (35%)
Scheme 39. Photochemical rearrangement of nitrile imine 5611
All the other photolyses reported with stable nitrile imines
unequivocally led to carbodiimides, except in the cases of Cphosphino-N-triphenylsilylnitrileimine 54 n[541and N-phosphino-C-thiophosphinoylnitrile imine 46 ni4”]which underwent nitrogen-nitrogen bond cleavage. The photolysis of nitrile imine
54n gave phosphinonitrile 116 in 85% yield along with several
541
G. Bertrand and C. Wentrup
REVlRlVS
+ -
R2P-CsN-N-SiPhg
54n
hV
+
R2P-CsN
coordination at nitrogen is known).[971Moreover, when the
reaction was monitored by N M R spectroscopy at -5O'C, an
intermediate tentatively assigned as isodiazirine 122 was detected. Many questions remain concerning the factors influencing
the photochemical reactions of nitrile imines.
116 ( 8 5 % )
S
s +hV
R~P-C~N-N-PRZ +
R$'-CsN
118 (86%)
+
L '
J
120 (35%)
7.2. Rearrangement of Nitrile Imines to Diazo Compounds
Scheme 40. Photochemical rearrangement of nitrile imines 5411and 4611
silylated products which were not isolated (Scheme 40). Irradiation of 4611at 300 nm led to thiophosphinoylnitrile 118 and to
phosphinonitrene 119 which dimerizes into the cyclodiphosphazene 120. It has already been shown that 119, generated by
photolysis of bis(diisopropy1amino)phosphanyl-azide,behaves
as "phosphonitrile" l19',[y41and dimerizes to give the fourmembered ring compound 120[951(Scheme 40).
The photochemical rearrangement of nitrile imines to carbodiimides seems to be rather general. As indicated in
Scheme 41, the reaction is possible for compounds with a wide
variety of substituents at the carbon and nitrogen termini,[", 58.63. 691
hv
+
The possible existence of a diazomethane-nitrile imine equilibrium was described in Section 2 [Eq. (I)]. Nitrile imine --t
diazoalkane rearrangements have been postulated to explain the
products obtained in the thermolysis of potential nitrile imine
precursors.[y8.y91 However, the nitrile imines were never observed, and apart from one case[981the resulting diazo compounds were also not stable under the experimental conditions.
Moreover, even the thermolysis of 5-aryltetrazole 124["] leading to the isolable diazo compound 126 may proceed by an
alternative mechanism that bypasses nitrile imine 12511 and involves instead the tautomeric 5 H-tetrazole 127 (Scheme 43).
Recent a b initio calculations support the latter mechanism.[' Ool
X-N=C=N-Y
0 [crn-l]
Y
SiiPrg
2200
B(NiPr2)z
2200
B(NiPr2)z
SiiPrg
2210
2200
B(NiPr2)z
2200
GeRR'R"
2165
GeRR 'R "
B(NiPr2)z
2 1 06
2252
Scheme 43. Formation of diazo derivatives from 5-aryltetrazoles.
Scheme 41. Photochemical rearrangement of nitrile imines into carbodiimides Wavenumhers
of IR absorption bands of the
carbodiimides are given in the
right column.
As mentioned above, the nitrile imine --t carbodiimide rearrangement was postulated to proceed through an 1 H-diazirine intermediate which leads to an imidoylnitrene by ring
opening.[421Interestingly, one piece of evidence for this mechanism has been reported.[96]Addition of tetrachloro-o-benzoquinone to nitrile imine 4611 led to compound 123, which was
isolated after workup (Scheme 42). This compound can be regarded as an imidoylnitrene stabilized by donor- acceptor interaction with an oxygen atom (the stabilization of nitrenes by
The first definitive proof for the nitrile imine --f diazoalkane
rearrangement was reported in 1988.150]Indeed, nitrile imine
4611,which was fully characterized (including single-crystal Xray diffraction), rearranged on heating in chloroform ( 5 5 "C,
6 h) to the stable phosphino(thiophosphinoy1)diazomethane
46d (Scheme 44). However, this rearrangement does not appear
x :s
E:PR2
x + R~B-CH-N-E
N2 5 3 6 ( 6 5 % )
5
R 2P- - S i Ph3
N2 55d
[
s
+ TCBQ
R~E-CEN-N-PRZ + R$-CIN-N;P\R~
46n
-78'c
l:$cl]
R : iPr2N
s
R: i P r 2 N
x :lp
2 5 T , 24 h
R~P-$.-Silfeg
N2 5 2 6 ( 6 0 % )
Scheme 44. Rearrangement of stable nitrile imines into diazo compounds
S
Rzh-C,=Y
p
-
S
Rz;-C=N
to be general. So far, only three other examples have been published, and all of them involve C-phosphorus-N-silyl-substituted nitrile i m i n e ~ . [ ~ ~ ]
F o r the sake of comprehensiveness, the formation of bis(diisopropy1amino)thiophosphinoyldiazomethane (127d) and
phosphane oxide 128 from hydrolysis of nitrile imine 46n should
ac14 acl,
,P\R~
~,500c
Nc ,h,2
123
122
(84%)
Scheme 42. Evidence for the rearrangement nitrile imine + isodiazirine
doylnitrene.
542
+ imi-
Angeb$ Chem Inr Ed Engi 1994, 33. 527-545
REVIEWS
Nitrtle Imines
be mentioned. This reaction takes place on silica gel and has
been rationalized in terms of a nitrile imine -+ diazoalkane reThe first step could be the hydrolysis of the
phosphorus-nitrogen bond leading to the N-unsubstituted nitrile imine 12911and the hydroxyphosphane 130, followed by
isomerization to 127d and 128, respectively (Scheme 45).
s
R ,i:
r
S
Ri!-$-H
+
N2 1276
Si02
+ -
- cEN-N-PR
(H2O)
S6n
8. Conclusion and Outlook
0
R2b-H
128 ( 8 3 % )
T
(82%)
.....
R iPr2N
Scheme 45 Hydrolysis of nitrile imine 4611
The factors governing the rearrangement of nitrile imines to
diazo compounds are not understood. At the most, it seems that
the larger the substitutent on the nitrogen atom, the slower the
rearrangement.
The reverse rearrangement of diazo compounds to nitrile
imines has been postulated to explain the formation of N-(thiophosphinoy1)pyrazole 134 from the photolysis of thiophosphinoyldiazomethane 127d in the presence of dimethyl acetylenedicarboxytate.f'Ol]Pyrazole 134 is the expected product of the
[2 + 31 cycloaddition of nitrile imine 133n and the alkyne, and
it could well be that the diazo -+nitrile imine rearrangement
proceeds via 131 and 132 (Scheme 46). In other words, this
rearrangement would be assisted by the presence of the phosphorus-sulfur double bond.
R :i P r 2 N
R ' : CO$fe
R;
133n
Scheme 46. Example of a diaro compound
-
'R,
134
(70%)
nitrile imine rearrangement.
This mechanistic hypothesis is corroborated by the formation
of carbodiimides 135 and 136 on photolysis of (trimethylsilyl)(thiophosphinoyl)diazomethane 53d and 1,3,4,2hs-thiadiazaphosphole 100a, respectiveIy.[loL1In fact, these reactions
could be examples of the rearrangement of a diazoalkane to a
nitrile imine to a carbodiimide (Scheme 47).
(ip r2N) #,
s
-$ - S i n q
L
b
N2
536
3,
s
2
(iP r 2N) 2P -N=C=N-S ine 3
135 (60%)
hV
1
s
( i P r 2N) 2P C-P ( N i P r 2 )2 --* ( i P r 2N)2P-N=C=N-P ( NiPr, ) 2
u
I1
N-N
i0Oa
Numerous questions remain concerning mechanisms of the
rearrangements of the various isomers of nitrile imine. One of
the most fundamental ones is whether they are intra- or intermolecular. Future work in our laboratories will address such
questions.
136 ( 8 0 % )
Scheme 47. Rearrangements ofdiazo compound 53d and thiadiazaphosphole lOOa
to carbodiirnide?.
Nitrile imines were only accessible as reactive intermediates
until a few years ago, and since then many stable compounds of
this type have been prepared. Some of them are solids, others
distillable liquids. Most of the stable nitrile imines have been
prepared by nonclassical routes, and these compounds are now
available in multigram quantities. Their stabilization is determined largely by steric interactions. For this reason, the isolable
nitrile imines tend to be less reactive. They are also less ambiphilic than their transient counterparts; that is, they are usually nucleophilic, with the exception of the positively charged
N-phosphonionitrile imine 7811,which is strongly electrophilic.
The question of concertedness in 1,3-dipolar cycloaddition
reactions was previously much debated. The addition of the
stable nitrile imine 46n to dimethyl maleate has been shown to
be nonconcerted. Examples of nonconcerted cycloaddition of
other 1,3-dipoles have been reported previously.
A striking feature of stable nitrile imines is the possibility of
reactions at the periphery of the C N N skeleton, which allow the
synthesis of other nitrile imines. Reactions involving the nitrile
imiiie moiety and a heteroatom substituent are also noteworthy.
X-ray crystal structures have now been obtained for several
stable nitrile imines. While all of these compounds possess nonplanar, allenic structures, electron-withdrawing substituents
promote a partial linearization in the direction of the propargylic structure. This is the case of the C,N-bis(bory1) compound
85 n and the N-phosphonio derivative 78n. These unusual structures have generated a renewed interest in high-level a b initio
calculations of nitrile imines. While calculations at the HartreeFock level indicate that the allenic and the propargylic forms
can coexist as discrete minima. incorporation of electron correlation with Mdler-Plesset perturbation theory leads to the prediction that only the allenic form should be stable. However,
substituents on the CNN framework can strongly modify the
geometry. The agreement between experimentally determined
structures and those calculated for model compounds is excellent. These results will prompt further high-level calculations of
structures and energies for nitrile imines.
The variability of structure between the allenic and propargylic extremes makes a rigorous interpretation of the infrared
spectra difficult. Both theoretical and experimental work is
needed here, in particular regarding the hitherto elusive C-arylnitrile imines.
Another area of interplay theory and experiment concerns the
rearrangement of nitrile imines to diazo compounds and to carbodiimides. Theory predicts a very high activation energy both
for the 1,3-hydrogen shift (by either concerted or homolytic
pathways) and for the ring closure to the 1 H-diazirine intermediate. However, there are several examples of such rearrangements taking place under mild conditions. Further studies of
their mechanisms are highly desirable.
543
G . Bertrand and C. Wentrup
REVIEWS
The story of nitrile imines clearly illustrates that heteroatoms
are wonderful tools for stabilizing highly reactive species. The
variety of easily available heteroatom substituents with disparate electronic effects allows a ready understanding of the
factors that influence the stability of the compounds. This simplifies the choice of the right organic substituents required for
the synthesis of purely organic. stable species.
It is now clear that nitrile imines are no longer laboratory
curiosities; they are easily available and storable, and their use
as building blocks for organic synthesis or for polymeric materials is anticipated.
We rvould like to thank the Centre National de la Recherche
Scientifique J C N R S ),fbr a Visitor's Fellowship ( t o C . W ) Lihicli
made thi.r collaboration possible. Thanks are clue also to our coicwkers A . Baceiredo, R. Riau, G. Sicard. M . Granier, E: Castan,
M . P. Artlztir, K. Horcliler ~ o i iLocqiieiig/iien, M . Soleilliavoup,
a i d G. kw,ziarii in Toulouse, and G. G. Qiao, R. Leung-Toung,
and M . W Wong in Bi-isbane, without ,i,lioni this work would
n e w r have conie to ,fi.uition. The financial support of' the experimental work in Toulo~sebj' the CNRS and Pierre Fubre MPdiccirnents is greutfiilly acknowledged.
Received. January 14. 1993 [A 932 IE]
German version: ,4ngeii,. Cheni 1994. 106. 549
[I] a) R. Huisgen. M. Seidel. J. Sauer, J. W. McFarland. G. Wallbillich, J Org.
Chen?. 1959, 24. 892: b) R . Huisgen, M. Seidel, G Wallbillich. H. Knupfer.
Ewuhedron 1962. 17, 3.
[2] R. Huisgen. Angeii~.Chem. 1963. 27. 604; Angew. Chiw?. Int. Ed En:/. 1963.
2. 565.
[3] R. Huisgen. Angciv. C/?rni.1963. 27. 741 ; Angcmi. Cheni. h i t . Ed. Engl. 1963,
2. 633.
[4] R. Huisgen. J. Org. Cliem. 1976, 41. 40.3.
[S] R. Huisgen in 1 ,3-Dipolur C'jdurid(litron C/iemi.vtry (Ed.: A. Padwa), Wlley.
New York. 1984. Chapter 1. p Iff.
[6] P. Caramella. P. Grunanger In 1,3-Dipoliir Cyclouddirion Cheniisrr). (Ed.: A.
Padwa). Wiley, New York, 1984, Chapter 3. p. 291 ff.
[7] R. N. Butler, F. L. Scott, Cheni I d (London) 1970, 1216: Yu. P. Kitae\, B. I
Buzykin, Ru.s.c. Cliern. Rev. iEngl. Trans/.) 1972, 41, 495: A. S . Shawali, C.
Parkanyi. J. Hererurycl. CIi~ni.1980. 17, 833.
[XI R. A. Firestone. J. Or:. Chert?.1968. 33. 2285; ihirl. 1972. 37, 2181; Tetrahedron 1977. 33. 3009
[9] P. Caramella, K. N. Houk. J Am. Clieni. Soc. 1976, 98, 6397: P. Caramella.
R. W. Gandour, J. A. Hall; C. G. Deville. K . N. Houk. ihid. 1977. 99. 385.
[lo] Cf also P. C. Hiberty. C. Leforestier, J Am. Cheni. Sac 1978. 100.2012; P. C.
Hiberty. C. Ohanessian. ;hid 1982, 104. 66; C. Thomson. C. Glidewell. J
Crnnput. Chrni. 1983. 4, I
[11] K. N. Houk, K. 'fdmaguchi in 1,3-Dipolur Cyclouddirion Chmiislr). (Ed.: A.
Padwa), Wile). New York. 1984. Chapter 13. p. 407ff.
1121 M. W. Wong. C. Wentrup. J. A m . Chetii. S o ( . 1993. 115. 7743.
(131 E. Muller. W Kreutzmann. J w r u s Liebigs Ann. Ciiem. 19.34. 512. 264.
[I41 E. Muller. P. Kistner. R. Beutler. W. Rundel. H. Suhr. B. Zeeh, Jiisru.r Lrehixs
Ann. C/icw. 1968, 713, 87.
[l5] For a definitive structure determination of isocyanamide H,N- N C see E.
Schifer. M. Winnewisser, Ber. Bunwnges. Phjs. Cliern. 1982. 86. 780.
1161 J:P. Anselme. J Choii. Edur. 1977. 54. 296.
a ) An IR band at 2169 cm- ' obtained on photolysis o f diazomethane at 4 K
was tentatively ascribed to HCNNH: J. F. Ogilvie. Pliotocliern. Photohiol.
1966. 9, 65, cf. J. F. Ogilvie, J Mol. Specrrosc. 1969. 3, 513. However. this
band was possibly due to carbodiimide H N = C = N H , which absorbs at 2097
and 2 1 6 9 c m ~ ' i n A r m a t r i x [ 1 7 b J .Diaromethaneitselfabsorbsat 2099cm-'
at 4 K and thus would mask the band at 2097 c m - ' ; b) S. T. King. J. H.
Strope, .I CIi~ni.P/ip\. 1971. 54. 1289. For the gas-phase IR spectrum of
H N = C = N H (2105 and ca. 2177cni-I) see also M . Birk. M. Winnewisser.
Chern. P h w Lrtt. 1986, 123. 386; c) For earlier suggestions of the formation
of formonitrile imine HCNNH in photolyses of diazomethane, see D. E.
Milligan, G C. Pimentel. J. Chem. Phps. 1958. 39, 1405: 1. M. Mills. H. W.
Thompson, Trunr. Furuduy So< 1954. 50, 1270. However. the presence ofany
tautomcr was discounted by G. B. Moore, G. C. Pimentel. J Chein. Phps.
1964. 40, 329, 342.
W. Beck. K. Fcldl, Angmv. C/?cnr.1966, 78. 746. A r i : ~Chiw
~
In!. Ed. Engl.
1966.5, 722: W. Beck. P. Swoboda. K. Feldl. R. S. Tobias. Cheni. Ber. 1971.
104. 533. For the history of fulminic acid and its structure. see M. Winnewisser. B. P. Winnewisser. Chem. Lisrp 1976, 70. 785.
[I91 R . Fusco. R. Justoni, Cu;:. Chini. Ilul. 1937.67,3; R. Fusco. Pyrfi-oles in The
Chemisrry o f H e r w o c l d i c Compounds, Vol. 22 (Ed.: A. Weissberger). Wiley.
New York 1967, pp. 1-176.
[20] R. Fusco. R. Romani, Gazr. Cliim. liul. 1946. 76, 439; 1948, 78, 342.
[21] a) E. Bamberger, J. Grob. Ber. Drsch. Chem. Ges. 1901, 34, 523; b) "The
sodium salt of the isonitro compound decomposes to nitrile and the molecular
residue C,H,-C-N=N-C,,H,.
which immediately polymerizes to the tetrazoline derivative".
[22] J. S. Clovis. W Fliege, R. Hulsgen. Cheni. Biv. 1983, 116. 3062.
[23] R . Huisgen. J. Sauer. M. Seidel, Chem. Ber. 1961. 94, 2503.
124) R. Huisgen. J. Sauer, M. Seidel, Ju.\tu;, Liehi:., Ann. Chm?. 1962. 654. 146.
1251 P. Scheiner. J. F. Dinda, Teirtihe(1ron 1970. 26, 2619.
1261 M. MBrky. H. Meier. A. Wunderli. H. Heimgartner, H. Schmid, H.-J. Hansen.
Hcli,. Chim. Acta 1978. 6 1 , 1477.
[27] M. Ruccia, N.Vivona. Chim. Ind. (Milun) 1966. 4K. 147.
[28] H . Meier, W. Heinzclmann. H. Heinigartiier. Chimiu 1980. 34, 506. See also
H. Meier, H. Heimgartner, H d r . Chini. Actn 1977. 6 0 , 3035.
129) C. Griindmann, K . Flory. Jnstirs Liebi,yr Ann. C/ii.n7. 1969, 721, 91.
[30] W. A. F. Gladstone. J Cheiir. Sac. Chern. Coniinun. 1969. 179; W. A. F. Gladstone. J. B. Aylward. R. 0. C. Norman, J. Chern. Sot.. C 1969, 2587
I . Jpn. 1971,44, 1667; ihid.
[31] a) Y . Huseya, A. Chinune. M. Ohta. Bull. C ~ J WSo<.
1972, 45. 3203: h) C. S. Angadiydvar, M. V. George. J Org. Chem. 1971, 36.
1589; c) M . Mirky, H.-J. Hnnsen, H . Schmrd. H d i , . Chiin. A o u 1971, 54.
1275.
1321 a) J. Saucr. K. K . Mayer, Trtruhedron Lt,rr. 1968. 32: b) see also G . Westphal.
D. Schlangstedt. Z. Clirnl. 1979, 19. 217.
1331 H. Reimlinger. J. J. M. Vanderwjalle. G. S. D. King. W. R. F. Lingier. R.
Merenyi. Cheni. Ber. 1970. 103. 1918.
[34] J. H. Boyer. P. S. Ellis. J Chein. Soc. Perkiii Furis 1 1979. 483: J. H. Boyer,
P. J. A. Frints. L. HefPuori,c/ Chmi. 1970, 7. 71
(351 J. W Lown, B. E Landberg. Cun. J Chern. 1975. 53, 3782.
1361 H. Meier. W. Hemzelmann. H Heimeartner. Cliirnin 1980. 34. 504.
N. H. Toubro. A. Holm. J Ani. Chem. Soc. 1980. 102. 2093.
a ) A. Padwa. P. H. J. Carlsen. J. Am. Chrm. Soc. 1976. 98. 2006: 1979, Y9.
1514; b) 3. Fisher. W. Steglich. Angeii-. Chem 1979. 91. 168; Angew. C h w .
Int. Ed. Engl. 1979. 18, 167: c) A. Padwa in 1,3-Dipolur Cyclouddifion Ckenii s t r y (Ed.: A. Padwa). Wile).. New York, 1984, Chapter 12, pp. 277-406.
d) A. Padwa, AnRew Cheni. 1976. 88, 131: Angew. Cheni. Inr. Ed. Engngl., 1976.
rup. A. Damerius, W. Reichen. J. Org. Cheni. 1978. 43. 2037.
C. Wentrup, J. Benedikt. J Org. Cheni. 1980. 45. 1407
W Reichen. H d s . Chini. Actu 1976. 59. 1636.
S. Fischer, C. Wentrup. J Ciiem. Soc. Chrm. Cornmim. 1980, 502.
C . Wentrup. S. Fischer, A. Maquestiau, R. Flammang. Angew. Chem. 1985.
97. 74; Angew ('heni. In!. Ed. Engl. 1985. 24. 56.
H. Bock, R. Dammel. S. Fischer, C. Wentrup. Tetrahedron Letr. 1987,28,617.
A. Baceiredo. A. Igau. G. Bertrand, M. J. Menu, Y Dartiguenave, J. J. Bonnet. J. Am. C h n . Soc. 1986, 108, 7868.
G. Sicard. A. Baceiredo. G. Bertrand. J A m . Chem. SOC.1988, 110, 2663.
For a review see, for example. E.C. Taylor, I. J. Turchi. Chem. Rev. 1979, 79,
181.
S. Patai', Tile C'htvtristri of Diuronium und Diuro Groups. Wiley, New York,
1978; M. Regitz. Diazoiilhune.r.Thieme. Stuttgart. 1977; M Regitz. G. Maas.
Diuzo Cortipoun& Proptvties und Spiithens, Academic Press, Orlando. FL,
USA, 1986.
M. Granier, A. Baceiredo, Y. Dartiguenave, M. J. Menu. G. Bertrand, J Am.
Clicni. Soc. 1990, 112, 6277.
M. Granier, A. Baceiredo. G Bertrand. Anget?. Chem. 1988, 100, 1397:
Anjieii-. Clieni. Inr. Ed. Engl. 1988, 27. 1350.
A. Baceiredo. M. Granier. F. Castan. J. M. Sotiropoulos, G. Sicard. G.
Bertrand. Phospliorrrs Sn/fur Silicon Relut. Elein. 1990. 49. 131
F. Castan. M. Granier. T. A. Straw, A. Baceiredo. K. B. Dillon, G. Bertrand,
Chmi. Ber. 1991, 124, 1739.
G. Bertrand. Hereroar. Chrn?. 1991. 2, 29.
F. Castan. A. Baceiredo. D. Bigg, G. Bertrand, J Org. Chem. 1991,56, 1801
M. J. Menu. M Dartiguenave, Y Dartiguenave. J. 1. Bonnet. G. Bertrand, A.
Baceiredo. J. Orgunornet Chem. 1989. 372. 201.
R. Reau. G. Veneziani. G. Bertrand, J Am. Chern. Soc. 1992, 114, 6059.
F. Castan. A. Baceiredo. G. Bertrand, Angeii Cherri 1989. /Of,1253: Angew.
Chmi. Ini. Ed. Engl. 1989, 28. 1250.
M. P. Arthur. H Goodwin, A. Baceiredo. K. B. Dillon. G. Bertrand.
Orgunortiefullics 1991. /O. 3205
M P. Arthur, A. Baceiredo. G . Bertrand. J A m . Chem Soc. 1991, 113, 5856.
J. Lorberth, S. H. Shin. H. Donath. S. Wocadlo, W. Massa, J Orgunornet.
Chrm. 1991,407,167: A Fadini, E. Glozbach, P. Krommes, J. Lorberth. ihid.
1978, 149. 297.
M. F. Lappert, J. Lorberth. J. S. Poland. J Cheni. Soc. A 1970, 2954.
R. Rean, G. Veneziani. F. Dahan. Angel<,.Chen?. 1992. 104. 481; Angew.
Chrm. h i . Ed. En:/. 1992. 31. 439.
Angpw. Chem. Inr Ed. Engl. 1994, 33, 527 -545
REVIEWS
Nitrile linines
(631 C.Leue. P. Jutzi, R. Reau. G. Bertrdnd. unpublished results.
[64] a) M Grdnier. A. Baceiredo, M. Nieger. G. Bertrand, Angeir. Cheni. 1990.
102. 1185: A n g ~ i i ~
Cheni.
.
I n r . Ed. Engl. 1990. 29. 1123; b) M. Grdnier, A.
Baceiredo. G . Bertrand. V Huch. M. Veith, Inorg. Chem. 1991. 30, 1161.
[65] F. Caslan. A Baceiredo, F, Dahan, N. Auner. G. Bertrand. J Chem SOC.
Cli~wi.Coniniun. 1992. 1274.
1661 M . P Arthur. A. Baceiredo. J. Fischer. A. De Clan. G. Bertrdnd. Sw7rhe.sis
1992. 43.
[67] N . Dubau. A Baceiredo. F. Dahan. G. Bertrand. unpublished results.
[68] R . H HouL B. A. Levi. W. J. Hehre. J. Compicr. Chem. 1982. 3. 234.
[69] G ( i Qiiio. C . Wentrup. unpublished results.
1701 R. A. Ewins. P. Lorencdk. K 7 Ha.C. Wentrup. J Am. Chon.Soc. 1991. 113.
7261. and references therein.
[71] M. Sluro. M. Yimaka\ca. T. Kubota. H. Koyama, J Chrm. Soc. Chem. Comm i n i . 1968. 1409: K . H. Bodenseh. K. Morgenstern. Z. ,Yrrrinforsth. A 1970.
25, 1 5 0 : M. Winnewisser. C/ieni. Phvs. Lerr. 1971. I f . 579.
[72] J. H. Tele.;. G. Maier, B. A. Hess, L. J. Schaad, M. Winnewisser. B. P. Winnewi\aer. C‘hfni. B w 1989. 122, 753.
[73] 0.
L. Chapman. .I.-P. LeRoux. J. An?. Chrm. Soc. 1978, 100. 282.
[74] E P Janulis. S. R. Wilson. A. 1. Arduengo, Tiiraliedroii Lrrt. 1984. 25. 405.
[75] A . J. Arduengo. private communication.
[76] C Wcnlrup. S. Fischer. H.-M. Berstermann, M. KuzdJ, H. Luerssen. K .
Burger. A i y o i ~ Ch(wi.
.
1986. 98, 99; Angeiv. Chem. h i . Ed. Engl. 1986. 25. 85.
[77] H -1. Hansen. H. Heiingartuer in 1,3-Dipolur C:wlnudtlirioii Chemi.rtri-(Ed.:
A . Padwa). Wiley. New York. 1984. Chapter 2, pp. 177-290.
1781 K . Horchler von Locquenghien. R Reau. G. Bertrand, ./. Chem. Sor. Chcm.
Ci~mniini.1991. 1192.
1791 F. de Sarlo. A. Brandi. A. Guarna. N Niccolai. J M a p . Rrson. 1982.50. 64.
[YO] R Sustniann. Pio-e Appl. C ’ h c w ~ 1974.
.
40. 569.
[Xi] A. Eckell. R. Huisgen. R. Sustmann. G. Wallbillich. D. Grashey. E. Spindler,
C ‘ l w n i . Bcr. 1967. 100. 2192.
[XZ] R . Huirgen. G. Mloston. E. Langhals. J Am. Chcni. Soc. 1986, 108. 6401: J.
Oi.,c. ( ‘ l i ~ w i 1986.
.
5/. 4085: R. Huisgen. E. Langhals, H. Noth, Tetrahedron
Leri. 1986. 27. 5475; G. Mloston. E. Langhals, R. Huisgen. ihitl. 1989, 30.
5373.
(831 H Quast. D. Regnat. E. M. Peters. K . Peters. H. G von Schnering, A n , ~ ~ i r .
Chc.ni. 1990. 102. 124: Angcw Chem In!. Ed. EngI. 1990, 29. 695.
[84] K N Houk. J Sims. C. R. Watts. L. J. Luskus, J. Am. Chpm. S o t . 1973. 95.
7301; K. N . Houk. J. Sims. R. E. Duke, R. W. Strozier. J. K. George. ibid.
1973. 95. 7187.
[SS] R. Huisgen. H. J. Slurni. M. Seidel, Chern Ber. 1961, 94. 1555.
[86] R. Huisgen. J. Sauer, M. Seidel, Chem. Ber. 1960, 9.3. 2885: H. Reimlinger,
J. J. M . Vandewalle, G. S. D. King, W. R. F. Lingier. R. Merenyi, h i d . 1970.
103. 1918; H. Reimlinger. .
I
.
J. M. Vandewalle, W. R. F. Lingier. ihid 1970.
103,1960; M . S. Gibson. Terrahrdron 1963. 19. 1587.
[87] R. Luckenbach, Dynamtc Slcveochemrslry o / Penlatoorhn<rrd Pliovp/iorir,
and Relured E/er77enis. Thierne. Stuttgart, 1973.
1881 For a review of cycloaddition reactions of tetracyanoethylene see. for example. A. J. Fatiadi, Sv7rhrsis 1987. 749.
1891 Very few examples of [ 3 + 21 and [4 + 21 cycloaddition reactions h i t h the
nitrile group of tetracyanoethylene have been reported: J. E. Franz. R. K.
Howe, H. K . Pearl, J Org. Cheni.1976. 41. 620; M. Brown, R. E. Benson,
ihid. 1966. 31. 3849; K . Burger, H. Schickaneder. M. Prinzel. Ln*higs Ann.
Chmi. 1976, 30; K . Burger, H. Goth. Angen.. Chem. 1980. 92. 836: A n p r u .
Cheni. In/. Ed. Engl. 1980. 19.810; K. Burger. H. Goth. W. Schontag, J. Firl,
rerrahcdron 1982. 38, 287.
[90] T. Facklam, 0. Wagner. H. Heydt. M . Regitz. Anxetv. Chem. 1990. lfJ2, 316;
Angew. Chem. Inr. Ed. Enpl. 1990. 29, 314.
[91] N. Dubau, A. Baceiredo. F. Dahan. G. Bertrdnd. unpublished results.
[92] a ) C. Thonison. C. Glidewell. L Cornput. Chen7. 1983.4.1 ; b) C. Guimon. S.
Khayar, F. Gracian. M. Begtrup, G. Pfister-GuillouLo. C’hcm. PA?.,. 1989,
138. 159: c) J. B. Moffat, J Mol. Srriccr. 1979. 52. 275; d ) R. Leung-Toting,
M. W Wong, C. Wentrup. unpublished results.
[93] For a review, see C. Wentrup, Adv. Heto.oti.d C’licm. 1981. 28. 231.
[94] G. Sicdrd, A. Baceiredo, G. Bertrand. J P. Majoral, A i i g e ~ C
. h ~ m1984.
.
Y6.
450; Angen. Chem. Int. Ed. Gig/. 1984. 23. 459.
[95] A. Baceiredo. G. Bertrdnd. J. P. Majoral. G . Sicard, J. Jaud. J Galy. .I A m .
Chem.SUC.1984, f06. 6088: A. Baceiredo, G. Bertrand. J. P. Majorul. F. El
Anba. G. Manuel, h i d . 1985, 107. 3945.
[96] M. Grdnier. A. Baceiredo. H. Grutzmacher. H. Pritzkow, G Bertrand.
Anprw. Chrm. 1990, 1112. 671; A ~ I ~ ~ P IChcJm.
I..
In/. E d Engl. 1990. 29,
659.
[97] P. A. S. Smith i n Azitles andNirrenes ( E d . : E. F. V. Scriven). Acadcmic Press.
London 1984.
[98] C. Wentrup. H e h . Chini. Arta 1978, 61. 1755.
[99] A. Padwa. T. Caruso, D. Plache, J. Chem. .Six. Chcin. Commnn. 1980. 1229.
A. Padba. T. Caruso. S. Nahm. J. Org. C / i m . 1980, 45, 4065: A. Padwa. T.
Caruso. S. Nahm, A. Rodriguez. J A m Chcm. SOC. 1982. 104. 2865.
[IOO] M . W. Wong. R . Leung-Touiig. C . Wentrup. J. Am. Cham. S O L 1993. 115.
2465.
(loll M . Soleilhavoup. A. Baceiredo, F. Dahan, G. Bertrand, f t r o r g Clwm 1992.
31. 1500.
Just what does a referee system achieve?
Regular readers of “Angewandte” are well aware that Dr. Hans-Dieter Daniel has researched this question
using Angeivundte Chemie as example. The preliminary results were presented in a communication on this
topic (Issue 2/93). Now the complete analysis by Daniel is available in book form: “The Guardians of
Science: Fairness and Reliability in Peer Review” by H.-D. Daniel, foreword by H. Noth (VCH, Weinheim,
1993, ISBN 3-527-29041-9, D M 78.-).
Документ
Категория
Без категории
Просмотров
4
Размер файла
1 830 Кб
Теги
nitrile, matrix, imine, compounds, characterization, stable
1/--страниц
Пожаловаться на содержимое документа