close

Вход

Забыли?

вход по аккаунту

?

Bis(trimethylsilyl)diimine Preparation Structure and Reactivity.

код для вставкиСкачать
2-Methoxy-4,5-methylenedioxyamphetamine(14 b) can also simulate ring C and/or ring B, whereas 3-methoxy-4,5methylenedioxyamphetamine (14a) can assume only a
ring B conformation, as is also confirmed by the lower
activity of the latter.
methoxy groups, which allow a ring B orland ring C conformation. The methyl group on C-4 prevents rapid
enzymatic degradation of this compound, and this may
help to potentiate the action.
7. Closing Remarks
H3c\
H3C,
0 f i C H 3
(0
H
O-.H-N
WCH3
b
0
\A/
f N
On examination of the formulas of 2,3-dimethoxy-4,5methylenedioxyamphetamineand 2,5-dimethoxy-3,4-methylenedioxyamphetamine, it can be predicted on the basis
of the above working hypothesis that the second compound
will have a stronger action than the first. This has been
confirmed by pharmacoIogica1 tests.
Finally, mention should also be made of 2,5-dimethoxy4-methylamphetamine (DOM), which has become very
popular in the USA as a psychodelic drug under the name
STP. The very intense psychotomimetic action of this
substance can be explained by the favorable position of the
Though no specific receptor localizations and points of
attack on the molecular level have been established as yet
for LSD and the other hallucinogenic drugs described here,
the above working hypothesis developed by Snyder and
Ri~helson"~'
offers a valuable contribution to the elucidation of the mode of action of these psychotomimetics. The
abuse of nutmeg as a narcotic drug has given fresh impetus
to pharmacological research in the testing of psychotropically active components, and this research has led through
structure-action relations to a model possessing unquestionable interest. However, nutmeg is now almost without
significance in practical medicine, and there is little likelihood, in view of its strong taste, that this well-known spice
will remain popular for long as a narcotic drug.
Received: October 20,1970 [A 818 IE]
German version: Angew. Chem. 83,379 (1971)
Translated by Express Translation Service, London
Bis(trimethylsily1)diimine :Preparation, Structure, and Reactivity[']
By Nils Wiberg"'
Based on investigations carried out in collaboration with Wan-Chul Joo,
Gerhard Schwenk, Wilfried Uhlenbrock, and Michael Veith
Dedicated to my father, Professor Egon Wiberg, on the occasion of his 70th birthday
The highly reactive compound bis(trimethylsily1)diimine (BSD), which was first prepared by
oxidation of lithium tris(trimethylsilyl)hydrazide, is light blue, sensitive to thermolysis and
hydrolysis, and ignites spontaneously in air. On the basis of electron transfer, acid-base, or freeradical reactions, it acts in particular as a (preparatively useful) redox system and as an agent
for the introduction of azo groups. Redox reactions lead by oxidation or reduction of the other
reactant through two oxidation stages to hydrazine derivatives or molecular nitrogen, and in the
case of electrochemical reduction, to BSD radical-anions.Azo-group transfers, on the other hand,
yield new inorganic azo compounds with no change in the oxidation state of the diimine group.
1. Introduction
When we succeeded, in 1968, in synthesizing bis(trimethylsilyl)diimine[zl
(CH,),Si-N=N--Si(CH,),
BSD
- ___
[*I Priv.-Doz. Dr. N. Wiberg
Institut fiir Anorganische Chemie der UniversitBt
8 Miinchen 2, Meiserstrasse 1 (Germany)
374
as the fist diimine completely substituted with silyl
groups131,the unexpectedly long-wave light absorption
and the resulting radiant blue color of the compound
immediately aroused our particular interest. The investigation of this substance, which has lost none of its original fascination, has proved worth while ; it has revealed
a varied, unusual, and interesting chemistry.
Various investigators had attempted to prepare a bis(sily1)diimine. The ability of carbon to fonn azo compounds (Z), which are often surprisingly stable (for
Angew. Chem. internat. Edit. / Vol. 10 (1971) 1 No. 6
example, azobenzene is stable up to 600°C[41),makes
the synthesis of "azosilanes" (2) a challenge in view of
Li,
RW-N,
+
SiR,
R,Si
Li
-+
,,N-N:
RSO,-N/
SiR,
\ SIR,
.
+ N,
+
(1)
R,Si-N=N-SiR,
the group relationship between carbon and silicon. A
possible preparation method was seen in the oxidation
of 1,2-bis(silyl)hydrazines(3).
x\
R,Si
/'
/N-N\
SiR,
8
-
(3)
Oxidation
-''
The arenesulfonyl azide, as in its reactions with CH-acidic
methylene derivatives"'], evidently adds first to the
silylhydrazide to form a nitrogen chain compound
R'SO,-N=N-N-N(SiR,)-N(SiR,),,
which can
decompose after migration of silyl groups to give the
final reaction products.
R,Si-N=N-SiR,
Instead of arenesulfonyl azides, arenesulfonyl chlorides
may be used as the oxidizing agents[1g! In this case,
however, the reaction no longer follows a single course:
(4)
However, several attempts to verify this reactionc5-'I
were initially unsuccessful because of failure to find an
oxidizing agent that could oxidize bis(sily1)hydrazines
(3) only to the stage of the readily oxidizable bis(sily1)diimines ( 4 ) without oxidizing these further to nitrogen''].
By the action of p-toluenesulfonyl azide on lithium tris(trimethylsily1)hydrazide (X =Li, Y = SiR, in (3)), we
were able to obtain BSD as a readily sublimable product
that crystallizes in .the form of needles and melts at
- 3 OC131.
Including BSD and a bis(bory1)diimine that is obtainable
from it["], there are now ten classes of isolable diimine
derivatives X-N=N-X
(counting diimine itself) in
which ligands X are linked to the azo group via maingroup elements having various atomic numbers:
R'S02
>-N:
R,Si
Li,N-Nf3R3
R'S02-C1
,
+
R,Si
SiR,
+ Licl
SiR,
(2)
SiR,
R'S0,Li
+ R,SiCl
+
R3Si-N= N- SiR,
Since BSD is also difficult to separate from trirnethylchlorosilane, the process is less suitable for the preparation of this diimine. However, it may offer advantages
in the synthesis of other diimines.
ff--N=N-H[ii]
yellow
>B-N=N-B=[lO]
yellow
?
3C-N=N-CG[4]
yellow-red
3Si-N=N-Si
pale blue
>N-N=N-N:[12]
colorless
[31
>0P-N=Nviolet
If one disregards diimine, BSD stands out from this
series of azo compounds because of its unusual properties. This compound, whose preparation, structure,
and reactivity will be surveyed below, reacts extremely
violently with itself and with oxygen, and can ignite
on contact with air. Since BSD is also readily hydrolyzed,
it is not easy to handle. Most of the operations described
in the next few sections must therefore be carried out
at low temperatures and in the absence of oxygen and
moisture.
2. Preparation
BSD can be prepared particularly easily by reaction of
an arenesulfonyl azide with lithium tris(trimethylsily1)hydrazide in ether at - 7 8 "C (R'= phenyl, p-t~lyl)[~]
:
Angew. Chem. internal. Edit. 1 Val. 10 (1971) 1No. 6
-O-N=N-Ocolorless
PO= [151
-0, S-N=N-S
deep yellow
[I31
F-N=N-F
colorless
0,- [i 61
(CI-N=N-F
yellow
[14,17]
[171)
Reaction (2) is evidently initiated by attack of the silylhydrazide on the electrophilic sites of the arenesulfonyl
chloride, i. e. the sulfur and the chlorine atom. The attack
on the sulfur atom leads to elimination of lithium
chloride and formation of tris(trimethylsily1)arenesulfonohydrazide, while the attack on the chlorine atom gives
lithium arenesulfinate and Nchlorotris(trimethy1silyl)hydrazine, which decomposes into trimethylchlorosilane
and BSD.
A third method for the preparation of BSD consists in
the thermolysis of N-lithiobis(trimethy1silyl)arylsulfonohydrazide at 180 "C:
SIR,
R'SO,
Li
\ N-N /
/
\
R'S0,Li
+ BSD
(3)
SiR,
375
3. Structure
in the ground state is shown in Fig. 1 (a)r251.
The energetically lower bonding n orbital and the two n orbitals,
as can be seen from the Figure, are each occupied by
two electrons, while the energetically higher antibonding n* orbital is empty. The energy difference between
the two n levels is more than 6 eV in the normal case.
The splitting of the two n levels, on the other hand, was a
subject of uncertainty for a long timerz6].According to
very recent photoelectron-spectroscopic investigation^[^^^,
the energy splitting, with a value of 3.32 eV in the case of
azomethane, is surprisingly large. The lower n orbital
thus lies below the bonding n orbital.
3.1. Molecular Geometry
Several structures are conceivable for BSD. The compound could contain a trans-angular (5), cis-angular (7),
or linear (6) SiNNSi skeleton. In addition to 1,2-bis(trimethylsilyl)diimine, the positional isomer 1,l-bis(trimethylsily1)diimine (8) must also be considered.
K=Np R 3
0 , 0 0 0
R3Si*N=N*SiR3
R3Si
.. ..
N=N,
R3Si
SiR3
R3Si,@ 2
N=N
R3Si'
E
1L
n+
3.2. Electronic Configuration
A typical term diagram of the inner n and n molecular
orbitals of the azo group
-f+fJ-
-6.09eV
- 7.69eV
A definite decision in favor of one of the three remaining
isomers ( S ) , (6), and (7) is possible only with reservations.
UV spectroscopic studiesfg1provide very definite evidence
against the linear form (cf. Section 3.2.1), and indicate
a relatively large SiNN angle (in comparison with CNN
angles in organic diimines). On the basis of the geometry
of trimethylsilyl isocyanate and isothiocyanate["], this
angle should be about 150" in BSD.
It is uncertain whether BSD is in the form ( 5 ) or (7).
Even at about -8O"C, the compound gives only one
signal in the 'H-NMR spectrum[g1.This could indicate
that one isomer is thermodynamically favored or, more
probably, that BSD consists of a mixture of cis- and
trans-isomeric molecules undergoing fast interconversion.
(Inversion on nitrogen, which generally occurs readily[231,
is reported to be additionally favored in silyl-nitrogen
compounds[241.)
+-/'
,/
The formally charged isomeric structure (8) is counterindicated by many physical properties of the compound
(e.g. volatility, solubility) and by IR"] and (indirectly)
ESR['O1 spectroscopic data (cf. Section 3.2.2). Moreover,
there is no chemical evidence of the existence of (8)
even in small equilibrium concentrations in the presence
of ( S ) , (6), and (7)["].
I
at
El
bl
dl
Fig. I. Term diagram of inner molecular orbitals: a) of the angular
azo group; b) of the linear azo group; c) of the azo system of
bis(tert-buty1)diimine [29]; d) of the azo system of BSD [29].
In addition to the n and n orbitals of the azo group,
it is also necessary to consider the (T and (T*molecular
orbitals.
3.2.1. The Color of BSD
One of the particularly remarkable properties of BSD is
its light blue color. The weak absorption that is responsible
for this color, a forbidden n++n* transition (cf. Fig. 1
(d)), occurs at 12750 cm-', i.e. in the invisible infrared
region of the spectrum. One "sees" only the short-wave
tail of the absorption band. Thus the wavelength of the
n + +n* transition of BSD is surprisingly long in comparison
with the longest-wavelength UV absorption maximum
of yellow to red organic azo compounds (cf. Fig. 1 (c)
and (d) and Table 1).
Table 1. UV absorptions (solvent: alkane) [9] and ionization energies ( I E ) [28] of tert-butyl- and trimethylsilylsubstituted diimines [29b] (R =CH,).
color
ZE (eV)
h>;;'(cm-')
h:;:* (cm-')
(E)
(E)
R,C-N=N-CR,
R,C-N=N-SiR,
R,Si-N=N-SIR,
pale yellow
7.49 f0.02
27200 (12)
5oooO (1800)
red
[a1
2~
pale blue
6.09 i 0.02
12750 ( 5 ) [b]
52100(1700) [c]
5
(9)
m (1100)
[a] Accurate determination has so far been prevented by experimental dificulties. I E is about 7 eV
[b] The strongly asymmetric band exhibits fine structure.
[c] Another band occurs at 40000 (260).
376
Angew. Chem. internat. Edit.
Vol. I0 (1971)1 No.6
It can be shown experimentally that the bathochromic
shift of the n, +x* absorption of about 2 x 7200 wave
numbers on replacement of the two "trimethylmethyl"
groups in bis(tert-buty1)diimine by the homologous trimethylsilyl groups is due both to a pronounced increase
in the n, energy level and to a slight decrease in the x*
energy level (cf. Fig. 1 (c) and (d)[z9a1).The mechanism
of this mutual approach of the levels, which has not yet
been fully elucidated, could be explained, e.g., by inductive and mesomeric effects[301and by repulsion effects
of bonding electrons[311of the elements E =C and Si
bonded to the azo group. It should also be observed that
widening of the ENN angle must also bring the n, and x*
energy levels closer together.
In the extreme case of the linear skeleton ?E-N=N-EG
(ENN angle = 180"), the n, and the x* levels are in fact
degenerate with each other (and the same is true of the
n- and x levels), so that the difference in the energy levels
becomes vanishingly small (cf. Fig. 1 (b)). Widening of
the ENN angle on replacement of tert-butyl groups
attached to the azo group by trimethylsilyl groups is
indicated by the observed decrease in the molar extinctionIZ6]of the n , m * absorption (cf. Table 1).
3.2.2. The Delocalization of Electrons in BSD
Both for the occupied x orbital and for the n, orbital of
BSD, there is a possibility of interaction with the empty d
orbitals of the silicon of suitable symmetry. Nothing is
known at present about the magnitude of this interaction,
which can be described e.g. by limiting resonance structures
such as
0 Q _.
[-R,Si <N=N-SiR,-R,Si--N=N
-SiR,-R,Si-N=N=SiR,
8 0
-3
On the whole, a weak but definite delocalization of the x
and n electrons is to be expected, since the ESR spectrum
of the BSD radical-anion[201,in the case of the unpaired
electron accommodated in the x* molecular orbital,
points to delocalization (though not very extensive) over
the entire SiNNSi skeleton.
The ESR line-splitting diagram of the BSD radical-anion
corresponds to an interaction in which the nitrogen nuclei
are equivalent[z01.The anion is thus not in the unsymmetrical 1,l form (8) but in the symmetrical 1,2-diimineform.
This result indicates that uncharged BSD is also in the
symmetrical form, and not in the equally conceivable
unsymmetrical form.
electron
BSDelectrochemical
redox reaction
electron
R -BSDf r e e radical
reactions
BSD-D
Lewis acid-base
reactions (redox
and substitution
reactions)
Scheme 1. Reactivity of bis(trimethylsilyl)diimine.
As has been confirmed experimentally, the donation or
acceptance of electrons can take place completely in an
electrochemical redox reaction or only partly in a Lewis
acid-base reaction (a redox reaction in the broader sense).
The primary reaction products obtained from BSD in
this way (cf. Scheme 1) are BSD cations and anions
(BSDf, BSD-) on the one hand and BSD-acceptor and
BSD-donor complexes (BSD-A, B S D t D) respectively on
the other, which can undergo further chemical reaction in
most cases. For example, BSD-acceptor complexes cannot
generally be isolated; one obtains only their further
reaction products, which can be formally described as
products of redox or substitution reactions of BSD (cf.
Scheme 2, Section 4.3).
It is also to be expected that radicals R which combine
the properties of an electron acceptor with those of a donor
will also add to BSD to form hydrazyl radicals (I-BSD).
According to the results that we have obtained so far,
this is in fact so. However, BSD is fairly stable toward
free-radical cleavage of the Si-N bond resulting in destruction of the BSD molecule (M),
which could conceivably occur from energetically excited BSD with simultaneous formation of molecular nitrogen. Thus no detectable
decomposition of BSD occurs on UV irradiation for
several hours, and BSD can be kept indefinitely in daylight.
The compound is also very stable toward heat, provided
that the molecules have no opportunity for collisions
leading to rapid decomposition.
The reactions of BSD are classified according to type in
Scheme 1, and numerous examples are given in the following sections.
4. Reactivity
BSD is extremely reactive, and reacts with practically
anything available. This striking reactivity of the azosilane
system can be deduced from the term diagram of the
compound (Fig. 1(d)). The inner molecular orbitals of BSD
that are important to chemical reactions, i. e. the highest
occupied n, molecular orbital and the lowest empty x*
molecular orbital, have very high and very low energies
Angew. Chem. internat. Edit. 1 Vol. 10 (1971)
respectively, unlike the corresponding orbitals of the
azoalkanes. This necessarily makes the compound particularly ready to give up and to receive electrons.
No. 6
4.1. Thermolysis Reactions
BSD decomposes below 0 "C. The thermolysis products,
the yields of which depend on the temperature, result from
disproportionation, dimerization, and free-radical reactions of BSD.
377
4.1.1. Disproportionation
The thermolysis products formed from BSD by disproportionation between - 35 and - 20 "C consist almost
exclusively of nitrogen and tetrakis(trimethylsily1)hydrazine ( 9 ) , which was synthesized for the first time in this
wayf31:
>-35oc
2 R,Si-N=N-SiR,
-----+
R,Si\
NZN
+
/N-(siR3
R3Si
(4)
SiR,
(9)
The azosilane BSD thus shows no similarity either in its
thermal stability or in its decomposition mechanism to the
very stable a z o a l k a n e ~ [ ~which
~ ] , decomposed in a first
order reaction, mainly into nitrogen and hydrocarbons:
which can generally be achieved in the case of diimine as
an uncatalyzed hydrogenation of apolar multiple bonds[' I],
is possible only with particularly reactive unsaturated
systems in the case of BSD, since the thermolysis of BSD
otherwise takes precedence. Thus unlike diimine["], BSD
does not react with "normal" carbon multiple bonds e.g.
of ethylene or cy~lohexene[~~].
Since the disproportionation of BSD (4) has a reaction
order greater than 1,the halflife of the thermolysis depends
on the concentration of BSD. In very dilute solution,
therefore, BSD can be kept for several days even at room
temperature, and at a pressure of 0.001 torr it can be
passed through pipes at 300°C without decomposing
since it is very stable toward a first order decomposition
with free-radical cleavage of SIN bonds corresponding to
the reaction exhibited by the azoalkanes (see above).
4.1.2. Dimerization
The decomposition of BSD (4) corresponds, however, to
the thermolysis of the parent diimine, which decomposes
into nitrogen and hydrazine even at very low temperatures" '1.
The thermolysis of BSD thus again supports the empirical
observation that from the chemical point of view, silyl
groups are less like organyl groups than like hydrogen.
The replacement of the hydrogen atoms of diimine by silyl
2 H-N=N-H
>-150%
NSN
+ H\ N-N(
H'
(6)
H
groups does not affect the stoichiometry of the decomposition of the diimine in any way, but merely leads to a considerable increase in the thermal stability of the molecule.
The similarity of the chemical properties of diimine and
BSD is not confined to thermolysis. Both diimine and BSD
can transfer their ligands from the azo group to a doublebond system other than their own in a reaction corresponding to the decomposition (4) or (6). For example, both
compounds quantitatively reduce both
and
azodicarboxylic esters" 1,341 even at low temperatures,
with hydrogenation and silylation respectively of the
apolar double bonds (X =H, SiR3):
If the thermolysis of BSD takes place above -3O"C,
tetrakis(trimethylsily1)tetrazene (10) is formed as well as
nitrogen and (9)[361in accordance with the overall
equation:
>-3OOC
2 R,Si-N=N-SiR,
There are three conceivable routes for the formation of
the tetrazene (10) by dimerization of BSD. These are:
1) Chain addition of two molecules of (R3Si),N=N;
2) cycloaddition of two molecules of R,Si-N=N-SiR,
followed by rearrangement of the resulting cyclotetrazane;
3) insertion of the azo groups of two molecules of BSD
into Si-N bonds in accordance with:
X-N=N-X
N=N
+
R,Si -N=N- SiR,
O=O
-+
ROOC\
,COOR
N-N,
x
Azoalkanes, on the other hand, do not react with oxygen
or azodicarboxylic esters, if one disregards the complicated
reactions that take place at higher temperatures. The
extremely ready oxidation (7) of BSD is the reason why
bis(trimethylsily1) peroxide can always be found among
the products of reactions of BSD (e.g. of the thermolysis).
According to our observations so far, the transfer of
ligands from the azo groups to double-bond systems,
37%
R3Si-N-N SiR3
d
I
II
I
R3Si N-N-SiR,
(10)
N-N f X - 0 - 0 - X
+ ROOC-N=N-COOR *
'
x
(9)
SiR,
The compound (lo), which is the first example of a silylnitrogen compound with a four-membered nitrogen chain,
is formed by a reaction path that has not hitherto been
observed for azo compounds, with dimerization of BSD.
The contribution of the dimerization of BSD (9) to the
total reaction increases with rising temperature, partly at
the expense of the disproportionation of BSD (4).
+
+
N -N=N-N\
R3Si
R,Si-N=N- SiR,
X-N=N-X
/SiR3
R3Si\
Since 1,I-bis(trimethylsily1)diimine has never been detected
despite attempts to intercept itLz1],the first reaction path
can probably be ruled out. The second reaction path is
opposed inter a h by the Woodward-Hoffmann rules on
the conservation of orbital symmetry[371,which forbid
supra,supra-cycloadditionsuch as is to be expected here.
We therefore consider the remaining reaction path (10)
which may be formally regarded as a substitution reaction
(cf. Section 4.4) as probable.
4.1.3. Free-Radical Reactions
With rising temperature, the thermolysis of BSD leads
not only to ( 9 ) and ( l o ) but also to increasing quantities
Angew. Chem. internnl. Edit. / Vol. I0 (1971) / N o . 6
of tris(trimethylsily1)hydrazine ( I I ) and bis(trimethy1sily1)amine (12) in accordance with the overall equations[36.381.
3 R,Si-N=N-SIR,
>-3OoC
___f
+2 H
NrN
+
R,Si\
,N-(
R,Si
2
> - 30%
(11)
SiR,
R,Si\
NaN + 2
2 R,Si-N=N-SiR,
H
N-H
(12)
R3Si
(12)
On the basis of equations (4),(9), (II), and (12), the silylnitrogen compounds are formed in yields of 21 % of (9),
20% of ( l o ) , 28% of ( l l ) , and 6% of (12) at 180°C
with benzene as the thermolysis medium. The hydrogen
atoms required for the formation of compounds (11) and
(12) come from the methyl groups attached to the Si in
BSD. The molecules that are damaged by removal of
hydrogen react further to form a large number of secondary
products that have not yet been investigated.
4.2. “Electrochemical” Redox Reactions
4.2.1. BSD as an Electron Acceptor and Donor
Figure 2 shows the energy levels of inner molecular
orbitals of the electron acceptor tetracyanoethylene
(TCNE) and of the electron donor tetrakis(dimethy1amino)ethylene (TDAE), together with the corresponding
orbitals of BSD. TCNE has an extremely low empty x*
orbital, and can therefore accept one or two electrons to
form monoanions or dianions (TCNE- , TCNEZ-)r391.
Conversely, the highest occupied a orbital of TDAE is
energetically extremely high ; TDAE can therefore give
up one or two electrons to form monocations or dications
(TDAE’, TDAEZ+)[401.
Since the empty a* orbital of
TCNE and the occupied 7c orbital of TDAE are comparable
in energy with the a* and n, orbitals of BSD, BSD should
possess the properties both of an electron acceptor and
of an electron donor and should be capable of accepting
or donating one or two electrons:
0
+e
R3Si-NLN-SiR3
-e
+e
R,Si-N=N-SiR,
-e
(I54
O Q
R,Si-N-N-SiR,
With toluene instead of benzene as the thermolysis medium,
the hydrogen abstraction takes place exclusively from
the C-bound methyl groups. In addition to the thermolysis
products (9) to (12), the main product found under these
conditions is a hydrazone (13) formed in accordance
with
4 R,Si-N=N-SiR, + H,C-R’
+
NZN
+
2 (11)
(13)
R,Si
+
\N-N= CH-R’
R,Si’
113)
( R = phenyl). Benzyltrimethylsilylhydrazines are also
formed as by-products. From the decrease in the yield of (9)
and the corresponding increase in the yield of ( 1 1 ) on
replacement of benzene by toluene as the solvent, it can
be deduced that the disproportionation of BSD (4) and
the free-radical reaction of BSD (11) involve a common
intermediate.
0
-e
R,Si-N=N-SiR,
(156)
The introduction of electrons into the BSD system or the
withdrawal of electrons from the system should also be
possible by reaction of BSD with electropositive metals or
with electronegative nonmetals of the main groups. As
can be seen from the following discussion, alkali and
E
1L
lev1
TCNE
ESO
TOAE
~
- 1.50
- 1.19
n’
n*
- - --
n++
--#-a
-609
-6.13
Fig. 2. Term diagram of the highest occupied and lowest unoccupied
molecular orbitals of tetracyanoethylene (TCNE), his(trimethylsily1)diimine (BSD), and tetrakis(dimethy1amino)ethylene (TDAE).
+ BSDThe tris(trimethylsi1yl)hydrazyl radical formed in accordance with equation (14)could then react further with the
trimethylsilyl radical, either by combining to form the
hydrazine (9) or by hydrogen abstraction to form the
hydrazine (11) (cf. also Section 4.3.2).
Angew. Chem. internat. Edit. J Vol. 10 (1971)
+e
BSD does in fact readily undergo electrochemical reduction to the radical-anion BSD-[301. On the other hand,
the expected high stability of the radical-cation BSD’ is
impressively shown by the appearance in the mass spectrum
of a molecular ion signal that is unusually intense for
silyl-nitrogen compounds.
,
’
It seems possible that the reactions (4) and (II), which
proceed independently of the non-radical dimerization of
BSD (9), are induced by the free-radical reaction (14a).
-e
R,Si-NgN-SiR,
No. 6
alkaline earth metals on the one hand and halogens and
chalcogens on the other react with BSD, whereas the
elements of groups 111, IV, and V, as well as hydrogen
(in the absence of catalysts) show no reactivity toward
BSD.
379
4.2.2. Reduction of BSD with Electropositive Metals[41]
When alkali metals in ethers are brought into contact
with BSD, the quintet signal characteristic of a radicalanion BSD- is immediately observed in the ESR spectrum[201.For example, lithium reacts quantitatively with
BSD in tetrahydrofuran (THF) at -78°C according to
(16a) to give Li[BSD]:
Li[BSD] evidently also forms a complex with itself
(suggested formula (16)), since only in this way can the
reaction order of 1found for the decomposition of Li[BSD]
be simply explained on the basis of an intramolecular
rearrange~nent‘~~].
Complex formation of this type is also
indicated by the observation that Li[BSD] is also formed
from LiJBSD] and BSD in THF.
Li,[BSDI
B S D+ L i
Li[BSD]
(16a)
As expected, BSD can also be reduced to the dianion by
alkali metals. Thus three days’ reaction of BSD with
lithium in diethyl ether at -30°C affords Li,[BSD] in
accord with (16b).
Unlike the BSD dianion, the resulting BSD radical-anions
are stable only at low temperatures, and decompose at
higher temperatures, depending on the metal M and the
solvent, with loss of nitrogen and formation of products
that include (R,Si),NM, (R,Si),N,M,,
MN,, and
(R,Si),N,. The course of the reaction is very complex, as
can be shown for the reaction of BSD with lithium in
THF, which proceeds quantitatively in accordance with
2 BSD + 2 Li
2 (R,Si),NLi
+ N,
(17)
This reaction is induced at -60°C by the formation of the
compound Li[BSD] as in eq. (16a). The latter compound
then decomposes in a first order reaction (with a half life
of z 1 hour at -45°C) into nitrogen and lithium bis(trimethylsily1)amide(14).
2 Li[BSD]
+
E
N
+
* 2 Li[BSDl
(20)
(16)
Conversely,by replacement of the solvent THF by pentane,
the compound (16) can be converted quantitatively into
Li,[BSD] and BSD; the indicated equilibrium (20) is displaced completely to the left because of the insolubility of
Li,[BSD] in pentane.
The concentration of free Li[BSD] in equilibrium with the
complex (15) is evidently so small that the decomposition
of the complex into nitrogen and (14) via Li[BSD]
2 (15)
--t
N,
+ 4 (14)
(21)
proceeds only very slowly; the first halflife of the decomposition (21) is about 1 week at -45°C. However, all the
initially formed Li[BSD] ultimately decomposes in
accordance with equation (I@,
with quantitative formation
of nitrogen and lithium bis(trimethylsily1)amide (14).
It should be mentioned in passing that Li[BSD] can also be
stabilized by complex formation with BSD in diethyl
ether.
Li
+ 3 BSD
{ Li[BSD]+ 2 BSD
-
Li(BSD),]+ BSD- (22)
(18)
However, one finds a total of only 50% conversion in
accordance with equation (18), from which it follows that
a reaction product (which must be ( 1 4 ) ) stabilizes unreacted Li[BSD] by “complex formation”.
+ (l4j
{LiBSD},
4.2.3. Oxidationof BSD with Electronegative Nonmetals[431
+ 2 LiN(SiR,),
(14)
Li[BSD]
+ BSD -pentane
THF
“Li[BSD]LiN(SiR,),”
(19)
(15)
Freshly prepared Li[BSD] accordingly does not decompose
at -45°C if an equimolar quantity of (14) is added to the
reaction solution.
Since the ESR spectrum of a solution of the complex (15)
points to three nitrogen atoms in the molecular complex,
two of which must be similarly bound, we consider the
following structural formula (15) to be possible :
The oxidizing power of nonmetals increases from left to
right and from bottom to top of the periodic table. In
agreement with this, BSD reacts with the halogens and
the more strongly oxidizing chalcogens oxygen, sulfur,
and selenium, but not with tellurium or the elements
of group V. However, the reactions with electronegative
nonmetals never led to BSD cations. This is understandable
if one considers the positive charge produced on the BSD
system on cation formation (cf. equation (15b)), which
greatly facilitates nucleophilic substitution of the anion
belonging to the BSD cation.
For example, tetrakis(dimethy1amino)ethylene (TDAE)
reacts quantitatively with halogen to form TDAE dihalideC4O1.
The reaction of BSD with halogen, on the other
hand, leads not to BSD dihalide but only to its formal
further reaction products, nitrogen and trimethylhalosilane.
I/N=N\
Li *
Li
\ 7
R3Si,
,SiR3
/“N\
Li
NAN\
TDAE + Hal,
BSD + Hal,
/N,
R3Si
SiR,
R,Si
Corresponding to reaction (24), oxygen (cf. reaction (7)),
S,, and red selenium oxidize BSD to nitrogen.
R3Si\
380
SiR3
/
\r”i
SiR,
+
+
TDAE” 2 Hal(BSD” 2 Hal-)
+
+
no further reaction
N=N+2R3Si-Hal
(23)
(24)
Angew. Chem. internat. Edit. j Vol. 10 (1971) 1 No. 6
BSD
+ lj8 E,
--t
E
N + R,Si-E-SiR,
substitution product is very often unstable, and decomposes with elimination of R,Si-S (cf. Scheme 2).
E=S, Se
The rate of the oxidation reaction decreases with increasing atomic weight of the chalcogen (the same is
true of the halogens), and the oxidation is increasingly
suppressed by the thermolysis of BSD in the order 0,
S, Se, Te:
Typical examples are provided by the reactions of BSD
with surfur dichloride and suIfur dioxide EX, (X=C1, 0)
to form nitrogen, sulfur, and trimethylchlorosilane or
nitrogen and bis(trimethylsily1) sulfoxylate at - 78 "C.
The formation of these products can be readily explained
on the basis of Scheme 2:
R,Si - X+E+X- SIR,
% oxidation
% thermolysis
100
0
2
98
47
53
0
100
(26)
R,Si - X-E - X - SiR,
The halogenation and chalcogenation of BSD naturally
need not follow the reaction (24). The reactions (24) and
(25) can probably be assigned to the redox reactions,
which will be discussed below.
The reaction sequence adduct formation, addition (adduct
formation, substitution, elimination) corresponds overall
to reduction (oxidation) of BSD through two oxidation
stages to give a hydrazine derivative (to give nitrogen):
4.3. Redox Reactions in the Broader Sense
4.3.1. BSD as a Lewis Acid and Lewis Base
The ability of BSD to act as an eitctron acceptor and
as an electron donor, which is dr.,r from the term diagram (Fig. 2), should also be reflected in Lewis acid
and Lewis base properties. Whereas on!! the complete
electron acceptance can be achieved preparatively in
the case of electrochemical redox reactions (formation
of BSD-, BSDZ-), the opposite is true of the acid-base
reactions, which proceed with partial electron transfer;
in other words, only reactions of BSD with Lewis acids
involving partial electron donation (formation of BSD+A)
have so far been observed, e.g. :he reactions of BSD with
lithium ions (cf. reaction (22)) and with trinitrobenzene
(h,,, of the CT absorption of BSD.nTNB : 21 100 cm-I),
which proceed with acid-base complex formation and
with EDA complex formation re~pectively[""~~"!
Such adducts of BSD with Lewis acids frequently cannot
be isolated since they are unstable and evidently decompose for the following reasons. Lewis acids S-B
generally contain basic centers B as well as acidic centers S. As a result of adduct formation between BSD and
the acid S-B
( = A in Scheme 1) the acidic centers
nitrogen and silicon of the BSD, which initially acts as a
base, are activated to such a degree that they react with
the (also activated) basic centers of the reactant S-B
that initially acted as the acid.
As a whole, therefore, the reaction corresponds to the
addition of S-B to the azo double bond or substitution
of a silyl group attached to the azo group by S. The
BSD
+
Redox reactions of the type (27) (and also substitution
reactions) are of preparative interest, since they often
proceed quantitatively. Reduction and oxidation reactions will therefore now be discussed in more detail
(substitution reactions in Section 4.4).
4.3.2. Reduction of BSD (BSD as Oxidizing Agent)
A typical addition reaction with reduction of BSD is
the reaction of BSD with organometallic compounds
(S-B=M-R
in Scheme 2), which leads in high yields
to metal G(trimethylsily1)organylhydrazides (M = Li,
MgBr, R' e.g. = methyl [cf. also reaction (52)], n-butyl,
phenyl)["].
BSD
+
M-R'
-
M,
/R'
N-N,
R,Si
SiR,
Metal hydrides (S-B = R,M-H) can evidently also add
to BSD, as was shown by the reaction of BSD with trimethylsilane to form tris(trirnethylsily1)hydrazine (11) [3s1.
BSD
+ R,Si-H
/H
R,Si,
N-N,
+
R,Si/
SiR,
(11)
However, reaction (29) probably does not follow Scheme 2,
since it proceeds in benzene only under the conditions
of the thermolysis of BSD, and is completely suppressed
S-B
1
adduct
formation
R,Si-B
+
NsN
+
S-SiR,
Scheme 2. Redox and substitution reactions ofbis(trimethylsily1)diimine.
Angew. Chem. internat. Edit./ Vol. 10 (1971) J N o . 6
38 1
in toluene by the free-radical thermolysis of BSD (Section
4.1.3)[451.Trimethylsilane probably adds to BSD by a freeradical chain reaction.
Another example of the reduction of BSD is the reaction
of BSD with “dichlorocarbene” formed as an intermediate
from chloroform.
BSD
+
“CC12”
-{
/T
R,Si -N-N -SiR,
form of the tetrachlorides, but the reduction of these
compounds by BSD must proceed via the dichlorides.
The reaction of BSD with disulfur dichloride is particularly
interesting. It leads quantitatively, in accordance with
2 BSD + S,C1,
2 N,
CH,CI,
}
+ 2 R,SiCI + R,Si-S-S-SiR,
(32)
(19)
(30)
The product is not the expected compound (17), but an
openchain isomer (18). The dichlorocarbene required
for reaction (30) can be formally obtained by reduction
of CCI, with BSD (cf. Section 4.3.3); the reaction of
BSD with carbon tetrachloride thus proceeds overall in
accordance with equation
In agreement with investigations on diimine” ’I, addition
of a diene (butadiene) to BSD (Diels-Alder reaction)
has not been observed.
4.3.3. Oxidation of BSD (BSD as a Reducing Agent)
We have so far found oxidations of BSD to be much more
common than reductions. Many nonmetals (cf. Section
4.2.3), nonmetal halides, and nonmetal chalcogenides,
as well as many other compounds, can oxidize BSD to
nitrogen. This is illustrated below by some reactions
of chlorides, oxides, and azides with BSD.
to bis(trimethylsi1yl)disulfane (19) if S,CI, is added
dropwise to a cooled solution of BSD. (Similarly, reaction
of BSD with trisulfur dichloride leads to bis(trimethy1sily1)trisulfane in a reaction resembling (32).) One possible
explanation of reaction (32) could be that, corresponding
to the expected reaction (26), S,CI, is first reduced by
BSD to disulfur, which could then react further with BSD,
in analogy with reaction (7) with oxygen, to form the
disulfane (19), which decomposes readily in accordance
with
+ 118 S ,
(19) -+ R,Si-S-SiR,
(33)
Diselenium dichloride probably also reacts in accordance
with equation (32), but instead of the expected bis(trimethylsilyl)diselane, one obtains bis(trimethy1silyl)selane
and selenium (cf. reaction (33)).
The reductions of some chlorine compounds ECl, of the
nonmetals of the first short period (E=N, 0, F) with
BSD, which have not yet been mentioned, differ from
the reactions discussed so far, since these compounds
contain positive chlorine. Instead of the central element
E, therefore, only the ligand chlorine can be reduced.
This is clear from a comparison of the reactions of
BSD with bis(organy1)chloroamine and bis(organy1)chlorophosphane in methylene chloride:
BSD + R2N-CI
BSD + 2 R,P-Cl
+ N,
+
N,
+ R,SiCI + R,N-SiR,
+ 2 R,SiCI + R,P-PR,
(34)
(35)
Reduction of
With the exception of silicon tetrachloride, all the chlorides
of elements E of main groups IV-VII react with BSD.
The nearer the elements of the chlorides ECl. are to the
right of the Periodic Table, the more extensively they are
reduced at -78 “C in methylene chloride:
~
(GeCI,)
SKI,
(PCI,)
AsCI,
SbCI,
SCI,
“SeCl,”
TeCI,
ClCl
BrCl
ICl
E”
E0
EO
E-’
~~
Reduction to:
It has also been found that in analogy with the reduction
of SnCl,, TiCl, is reduced to TiCI, by BSD, whereas
the formally expected GeCl, is not obtained, since it
evidently reacts further with BSD (cf. reaction (31)).
PCl, is not reduced to elemental phosphorus by BSD.
On the other hand, SCl, can be reduced to the oxidation
state -11 of sulfur at “higher” temperatures (O’C, cf.
reaction (25)). Selenium and tellurium were used in the
382
(Concerning reaction ( 3 9 , cf. Section 4.4; for the reaction of BSD with CCI,, cf. reaction (31).)
Reduction of Oxides1461
Whereas the o bonds of the nonmetal chlorides readily
undergo reductive cleavage by BSD, this agent preferentially attacks only n bonds in the nonmetal oxides,
so that the choice of possible reaction partners for BSD
is smaller in the case of the oxides than in the case of
the chlorides. Reduction reactions have so far been
observed only for carbon, nitrogen, “oxygen”, and sulfur oxides (carbon dioxide does not react with BSD).
Only the reactions of BSD with sulfur oxides will be
discussed below.
Suljiur dioxide, as was mentioned earlier (cf. reaction
(26)), is reduced to bis(trimethylsily1) sulfoxylate (20)
by BSD at - 78 “C [471.
BSD
+ SO,
+
N,
+ R,SiO-S-OSiR,
(36)
(20)
Angew. Chem. infernat. Edit.
VoI. 10 (1971) / No. 6
The very unstable sulfoxylate (20), which was obtained
in this way for the first time, decomposes above - 40 “C
into sulfur dioxide, sulfur, and hexamethyldisiloxane. In
the presence of sulfur trioxide, this decomposition takes
place even at -78”C, and in this case, as expected,
hexamethyldisiloxane is obtained as the SO, adduct,
bis(trimethylsily1)sulfate (21)1481.
lZO)- 112 so,,- 112 s’
% R,SiO-S-OSiR,
BSD + Ph-N=O
Ph
+
N,
+ R,Si-N-0-SIR,
(37)
R,SiO
0
BSD + 2 Ph,C=O
(21)
+
N,
OSiR3
i t
+ Ph,C-CPh,
The sulfoxylate (20) is also formed from BSD and sulfury1 chloride, since BSD probably first reduces the
chloride SO,Cl, to SO, (cf. above),
BSD + SO,CI,
+
N,
+ 2 R,SiCI + SO,
(38)
which then reacts further in accordance with equation
(36).
Sulfur trioxide is also reduced by BSD. The reaction
product to be expected here in accordance with reaction
(36), i. e. bis(trimethylsily1) sulfite
BSD + SO3 + N,
0
+ R,SiO-S-OSiR,
(39)
(22)
cannot be isolated, since it appears to be thermally
unstable and reacts rapidly with further sulfur trioxide
in a reaction similar to (37) to form sulfur dioxide
and silyl sulfate (21).
Since SO, can react first with BSD and then with SO,
in accordance with equations (36) and (37), the overall
reaction of BSD with sulfur trioxide ultimately corresponds to the formation of nitrogen, sulfur, and bis(trimethylsilyl) sulfate (21).
3 B S D + 4 S 0 3 + 3 N 2 + S + 3 (21)
(41)
Suljiir monoxide, which should occur as an intermediate
in the reaction of BSD with thionyl chloride (cf. reaction (38)), should also react with BSD in the same way
as oxygen (reaction (7)) :
Reduction of A z i d e ~ I ~ ~ l
Finally, the reactions with azides to form amines also
merit attention among the oxidations of BSD:
BSD + X-N,
2 N, + R,Si-S-0-SiR,
(42)
+
2 N,
+ X-N(SiR,),
(45)
Since BSD reacts only with electrophilic azides XN,
below 0°C (X=e.g. acyl, p-t~luenesulfonyl[~~~
(RO),Sit511,
but not phenyl, R,Si) and the rate of the reaction (45)
is increased by polar reaction media, (45) is probably
induced by electrophilic attack of the azide on the azo
system (formation of the adduct (24)) in accordance
with the general reaction Scheme 2. (24) could then
readily change into the observed products by migration
of silyl groups.
X-N=N-N
SiR,
t
R,Si-N=N-SiR,
-D
+
X-(
2 N=N
(46)
SiR,
(24)
Hydrazoic acid also reacts with BSD in accordance with
equation (45) (X =H) ; in this special case, however, the
reaction could also be initiated by attack on BSD by a
proton (cf. Section 4.4).
4.4. Substitution Reactions
4.4.1. BSD as “Azo-ation” Agent
Corresponding to the general Scheme 2 of the reaction
of BSD with Lewis acids S-B, BSD substitution reactions proceed in accordance with equations (47) and
(48).
R,Si-N=N-SiR,
S-N=N-SiR,
BSD + “ S O
(444
or with benzophenone:
0
(R,Si),O
Like SCl,, therefore, SO is reduced to sulfur. The fundamental possibility of reaction of element-oxygen groups
with BSD in the sense of reaction (42) is strongly indicated by the reaction of BSD with nitrosobenzene:
+ S-B
+ S-B
+
R,Si-B
+ S-N=N-SiR,
(47)
(25)
+
R,Si-B
+ S-N=N-S
(48)
(26)
(23)
However, the reaction of thionyl chloride with BSD leads
not to the compound (23), but to products that include
sulfur and hexamethyldisiloxane (formally the decomposition products of the compound (23)) in accordance with
2 BSD + SOCI,
+ 2 N,
+ 2R,SiCI + l j 8 S , + (R,Si),O
Angew. Chem. internat. Edit. 1 Val. 10 (1971) / N o . 6
(43)
Such reactions thus correspond to transfer of the azo
group to S; BSD thus acts as an “azo-ation agent” for S,
and could therefore be used for the preparation of
inorganic azo compounds (25) and (26). However, the
substituted diimines expected according to equations
(47) and (48) are frequently not obtained, since they are
often unstable and e.g. lose nitrogen (cf. Section 4.3) or
react further with BSD (cf. Sections 4.4.2 and 4.4.3).
383
Thus the protolysis of BSD (S-B in Scheme 2 = H - X ;
X =e.g. OH, OR, OSiR,, N H , , NHR, N,, C1, S0,H)
leads not to diimine, which is probably formed as an
intermediate in accordance with
Insertion of the carbonyl group of acetone
(S-B = R,C=O) into the Si-N bond of BSD also occurs
in the sense of reactions (47) and (48):
R
BSD + 2 H-X
-t
2 R,Si-X
+ H-N=N-H
(49)
+R,C=O R,Si--N=N-C-OSiR,
R,Si-N=N-SiR,
blue
R
red
but to nitrogen and hydrazine (in the form of its H X
adduct) together with other products (cf. Section 4.4.3).
The reaction of BSD with methyllithium (S-B = Li-R),
which proceeds with simultaneous addition (Section 4.3.2)
and substitution, leads not to the expected substitution
product lithium trimethylsilyldiimide, which is undoubtedly formed as an intermediate
BSD + Li-R
--t
+ Li-N=N-SiR,
R,Si-R
+ O=CR,
>
(55)
R
R
R,SiO-C-N=N-C-OSiR,
R
R
yellow
The reactions of BSD with organylchlorophosphanes
and with water will now be discussed as examples of BSD
substitution reactions in more detail.
(50)
(27)
4.4.2. Reaction of BSD with Organyl~hlorophosphaned~~~
but to lithium azide and lithium bis(trimethylsily1)amide.
Both compounds may be formally regarded as products
formed by further reaction of the azo compound (27)['9].
On consideration of reaction (28) and the thermolysis
and hydrolysis reactions of silylated diimines summarized
in Scheme 3 (cf. Section 4.4.3), it seems particularly
probable to us that the diimide (27) adds to BSD to
form a tetrazene compound, which then decomposes:
BSD
+
SiR,
Li
(27)
\N-N=N-N:
+
R,Si'
-+
(R,Si),NLi
SiR,
(51)
+
N=N=N-SiR,
According to our results[521,the trimethylsilyl azide
formed would react rapidly with methyllithium to give
tetramethylsilane and lithium azide, so that including
equations (50) and (51), the reaction course found experimentally can be represented by :
2 BSD + 2 LiR
--t
2 R,Si
+ LiN, + (R,Si),NLi
(52)
In the reaction of BSD with bis(organyl)chlorophosphanes
(S-B=R',P-Cl),
instead of the azo compound (28)
that is expected according to equation (47), and which,
according to experimental results, is probably formed
as a reaction intermediate:
BSD -tR;P-CI
+
R,Si-CI
+ R;P-N=N-SiR3
(53)
The reactions of BSD with organylchlorophosphanes are
sometimes rather complicated. The ease with which they
proceed increases with the degree of chlorination of the
phosphorus, i.e. with increasing acidity of the reactant.
Thus triphenylphosphane does not react with BSD ; diphenylchlorophosphane reacts at - 30 "C, phenyldichlorophosphane reacts slowly at - 78 "C, and phosphorus
trichloride is violently attacked by BSD at -78°C.
The nature of the reaction products is strongly dependent
on the reaction medium as can be illustrated for the
reaction of bis(organyI)chlorophosphanes with BSD.
With methylene chloride as the solvent, bis(organy1)chlorophosphanes react to form tetrakis(organy1)diphosphanes
BSD + 2 RkP-CI
CH CI
3N,
+ 2 R,Si-CI + R;P-PR;
(56)
The reaction proceeds in accordance with equation (53)
via a phosphanoazo compound (28), the formation of
which can be correlated with a deep green-blue color
that appears during the reaction. According to our investigations so far, (28) decomposes with liberation of
nitrogen.
Rip-N=N-SiR,
--t
N-N
+ Rip-SIR,
(574
(281
The bis(organy1)trimethyIsilylphosphane (which can be
isolated in the case of R'= mesityl) finally reacts with
bis(organy1)chlorophosphane as follows r521.
(28)
+ Cl-PR;
RkP-SiR,
On the other hand, the reaction of BSD with diphenylboron chloride (S-B = Ph,B--CI) proceeds as expected
in accordance with equations (47) and (48).
With diethyl ether as the solvent, bis(organy1)chlorophosphanes react much more slowly than in methylene chloride.
The product obtained on reaction of BSD with diphenylchlorophosphane is tris(trimethylsily1)diphenylphosphamidine (30) which can then react further with an excess of
BSD + 2 Ph,B-CI
+
2 R,Si-CI
+ Ph,B-N=N-BPh,
+
R,Si-CI
+ RbP-PR,
one again finds only its further reaction products (cf.
Section 4.4.2).
(57W
(54)
(291
This reaction allowed the preparation of the first pure
azoborane (29)
384
Angew. Chem. internat. Edit. 1 Vol. 10 (1971) 1 No. 6
which reacts further with water even during the primary
hydrolysis, some of these compounds, i.e. ( I I ) , (36),
and (12), are slowly hydrolyzed to hydrazine, ammonia,
and ammonium azide with formation of (34) and (35)
(reaction time for the secondary hydrolysis: a few weeks).
diphenylchlorophosphane (e.g. to form [Ph,P-Ph,P=N8
PPh,=N-PPh,-PPh,]Cl*).
This reaction leads surprisingly to a compound with isolated nitrogen atoms.
The diimine group must therefore have been split during
the formation of (30).
We assume that the reaction (58) again proceeds first via
I
the phosphanoazo compound (28) in accordance with
equation (53) (R = phenyl), and that (28) then undergoes
"1,3-dipolar" addition with BSD, as in related c a ~ e s [ ~ ~ * ~ ~ ] ,
with formation of the adduct (31) ;
PhzP-N-N-SiR3
+
+
?-FSiR3
PhZP,
R3Si -N=N-SiR,
(59)
r"'SiR3
R3Si
(3I)
j
I
Reactions of BSD with organyldichlorophosphanes or
even with phosphorus trichloride are even more complicated than those with bis(organy1)chlorophosphanes. This is
connected with the greater number of reactive chlorine
atoms. For example, the reaction of BSD with mesityldichlorophosphane leads inter alia to the phosphane (32) :
The formation of (32) can be readily explained on the
basis of the earlier discussions. Mesityldichlorophosphane
first undergoes a reaction corresponding to the sequence
of equations (53), (56) to give Mes(R,Si)PCl, which
reacts further with BSD in accordance with the sequence
(53), (59) to give compound (33). Like (31), compound
(33) can then decompose to form a phosphamidine (32a),
which would rearrange into the final product (32).
Mes,
//"N/SiR3
-
,p\
J
-N2
R3Si vHN\si~,
R3Si
(33)
Mes, #N(SiR3)
.p\
R3Si
N(SiR3),
(324
4.4.3. Reaction of BSD witb Water[351
When BSD is brought into contact with water at 20-40°C,
the reaction products include tetrakis(trimethylsily1)hydrazine (9), tetrakis(trimethylsily1)tetrazene ( l o ) , hydrazine,
ammonia, and ammonium azide, as well as nitrogen, trimethylsilanol R,SiOH (34),and the condensation product
of the latter (R,Si),O (35). A detailed investigation of the
hydrolysis of BSD showed that the reaction of BSD with
water takes place in two main reaction steps, which differ
greatly in their rates. Several fast reactions first lead to the
products shown in bold type in Scheme 3, which can be
detected as such (reaction time for the primary hydrolysis:
a few hours). Apart from bis(trimethylsily1)hydrazine(37),
Angew. Chem. internat. Edit. / Val. 10 (1971) / No. 6
slow'.'
Y
S
H,
I
S
The compound (31) could decompose, after migration
of two silyl groups to phosphorus-bound nitrogen atoms,
into the final products, i. e. nitrogen and (30).
lherrnolvsis
x'
N-N-N-N(
138/
H
x
x\
7 NH+N=N=NH
.
x$,?/
V
Scheme 3. Hydrolysis of bis(trirnethylsily1)diimine
[a] Hydrolysis leads to 1,l-bis(trimethylsi1yl)hydrazine and tetrazene.
The following observations should be added in connection
with Scheme 3. (9) and (10) are formed by thermal disproportionation and dimerization of BSD with itself (Section
4.1). The formation of (11) and (36) can also be explained
by corresponding reactions of BSD with mono(trimethy1sily1)diimine (MSD)[551.It may therefore be assumed that
the hydrolysis of BSD (as expected) proceeds via MSD as
an intermediate. MSD then reacts on the one hand
with BSD by disproportionation and dimerization to
give nitrogen, ( I l ) , and (36) (predominating principal
reaction), and on the other with water to give diimine.
The relationship of the reactions of BSD with BSD and of
BSD with MSD is particularly clearly shown by a feature
of the reaction (cf. Section 4.1.2), i.e. the increase in yield
both of tetrazene (10) and of (36) with rising reaction
temperature. The decomposition of diimines by two pathways (disproportionation, dimerization), which was first
observed in the thermolysis of BSD, is thus evidently of
more general importance.
Whereas (9) and (10) react only unmeasurably slowly
with water[561,(11) and (36) are hydrolyzed slowly but
measurably to hydrazine and ammonium azide respectively
(secondary hydrolysis). The unexpected formation of
ammonium azide can be explained as follows: The silyltetrazene (36) is hydrolyzed to tetrazene, which decomposes
in turn, with migration of hydrogen, into ammonia and
hydrazoic acid (= ammonium a~ide)'~~.''].
The fundamental possibility of such a decomposition is shown e.g. by
the thermolysis of (36), which leads quantitatively to (12)
and trimethylsilyl azide above 1oO°C[351.
Contrary to expectation, (10) does not decompose in a
similar manner into tris(trimethylsily1)amine and trimethylsilyl azide, but gives (12) and nitrogen above 150"C.
This free-radical decomposition, which corresponds to the
385
thermolysis of organic tetrazenes1”I, is probably due to
the fact that decomposition with migration of silyl groups
is impossible on steric grounds, as was shown by examination of the Stuart model of (10). The tetrazene (38),
which has not yet been isolated as such, should decompose
in the same way as the tetrazene (36), though more rapidly
because of the comparatively lower steric hindrance.
Since the compounds (11) and (36) undergo only very
slow hydrolytic decomposition and no thermolysis under
the reaction conditions, neither of these compounds can
be the source of the small quantities of hydrazine and of
ammonium azide formed in the primary hydrolysis.
However, initially formed hydrazine and ammonium
azide could be further reaction products of the doubly
silylated hydrazine (37) and tetrazene (38). According to
the above arguments, both compounds should be formed
from diimine (the final product of the hydrolysis of BSD
by the “direct” route) and BSD with disproportionation
and dimerization. It has in fact been possible to show
the probable presence, as a reaction intermediate at the
beginning of the hydrolysis of BSD, of the 1,2-isomer (37)
of bis(trimethylsilyl)hydrazine,which is itself sensitive to
watert5*’. On the other hand, attempts to find bis(trimethylsily1)tetrazene (38) have so far been unsuccessful.
The hydrolysis (and possibly also the thermolysis or a
reaction with BSD) of this tetrazene is evidently so fast
that only its decomposition products are found.
The compounds (37) and (38) could be formed from two
molecules of MSD. However, since the concentration of
the very reactive MSD probably remains small during the
hydrolysis, this reaction probably makes only an insignificant contribution to the process as a whole. For the same
reasons, reaction of diimine with MSD or diimine is also
less likely[591.
On the whole, therefore, the reaction of BSD with water
is a complicated system of successive and simultaneous
hydrolysis, disproportionation, nitrogen chain degradation, and nitrogen chain formation reactions. The situation
is further complicated by the fact that ammonium azide
and hydrazoic acid can also react with BSD in protolysis
(Section 4.4.1) or redox reactions (Section 4.3.3). The
latter reaction is probably responsible in particular for
the formation of ammonia (via (12) and its hydrolysis).
5. Outlook
It can be seen from this report that the reactive azosilane
BSD has diverse and unusual chemical capabilities as a
redox system and as an acid-base system, which have
presented us, and will probably continue to present us,
with many unexpected reactions (consider, e.g., BSD as a
complex ligand). The observation that the chemical
behavior of BSD is similar to that of diimine also deserves
special interest in this connection. It appears that the
change in properties on transition from nitrogen-hydrogen
compounds to their silyl derivatives is only slight, so that
the latter offer “thermally stable” mode1 compounds for
the investigation of the reactions of “thermally labile”
nitrogen-hydrogen compounds.
386
Thus, to give another example, tetrazene (N4H4) could
conceivably decompose into :
1) Hydrazine and nitrogen (+N,H,+N,);
2) Diimine (-2 N,H,);
3) Ammonia and hydrazoic acid (+NH3 +HN,).
The thermolysis and hydrolysis of silyltetrazenes (Section
4.4) now point to the third possibility, and oppose the
thermodynamically favored first possibility, whereas the
second reaction, according to results obtained so far
(Sections 4.1 and 4.4) can probably proceed only in the
reverse direction.
Our assumption of a close chemical relationship between
silyl-substitutedand unsubstituted nitrogen-hydrogencompounds offers encouragement to prepare and investigate
other silylnitrogen-hydrogen compounds, which should
also be ideal starting materials for the synthesis of “naked”
nitrogen-hydrogen compounds. Since feasible methods
for the preparation of such silyl derivatives, according to
preliminary results, again lead via BSD, our fascinating
“BSD adventure” has certainly not yet reached the limit
of its possibilities.
I am particularly grateful to the Deutsche Forschungsgemeinschaji and to the Fonds der Chemischen Industrie
for generous financial support of our investigations, as well
as to Dr. W.-Ch. Joo, Dr. G . Schwenk, Dip1.-Chem.
W . Uhlenbrock, and DipLChem. M . Veith for their
untiring and inspired cooperation.
Received: March 16,1971 [A 819 IE]
German version: Angew. Chem. 83,379 (1971)
Translated by Express Translation Service, London
[l] Compounds of silicon, Part 13.-Part 12: [41]. Also Part 7 of
derivatives of diimine.-Part 6: [41]. Presented in part as lectures:
2nd Int. Sympos. on OrganosiliconChemistry in Bordeaux (Juli 9- 12,
1968); Chemiedozententagung Koln (April 7, 1970); GDCh Berlin
branch (February 8, 1971); Serbian Chemical Society, Belgrade
(May 25, 1971); colloquia at several universities and polytechnics.
[2] Abbreviated hereafter to BSD; R in all formulas is CH,.
[3] N . Wiberg, W.-Ch. Joo, and W . Uhlenbrock, Angew. Chem. 80,
661 (1968); Angew. Chem. internat. Edit. 7, 640 (1968).
[4] Houben-Weyl: Methoden der Organischen Chemie 10/3, 219
(1965).
[5] H . Bock, Z. Naturforsch. 176, 423 (1962).
[6] U. Wannagat and C. Kriiger, Z. Anorg. Allg. Chem. 326, 288,
296, 304 (1964).
[7] R. West, personal communication.
[8) On the other hand, a series of oxidizing agents can be found [6, 91
for the preparation of organylsilyldiimines, which are more resistant to
oxidation, and which were described by Wannagat and Kriiger [6] as the
first azo compounds with a silyl group attached to nitrogen.
[9] N . Wiberg and M . Veith, unpublished.
[lo] N . Wiberg and G . Schwenk, Angew. Chem. 81, 745 (1969);
Angew. Chem. internat. Edit. 8, 755 (1969).
Ill]S. Hiinig, H.R. Miiller, and W . Thier, Angew. Chem. 77, 368
(1965); Angew. Chem. internat. Edit. 4,271 (1965).
1121 Houben-Weyl: Methoden der organischen Chemie 10/2, 828
(1967).
[I31 H . Kiefer and T.G. Traylor, Tetrahedron Lett. 1966, 6163;
D. J. Millen, C.N . Polydoropoulos, and D. Watson, J. Chem. SOC.1960,
687; J . Goubeau and K . Laitenberger, Z. Anorg. Allg. Chem. 320, 78
(1963); W . Beck, H. Engelmann, and H . S . Smedal, ibid. 357,134 (1968).
[14] C .J. Hofmann and R . G. Neville, Chem. Rev. 62, 5 (1962).
I151 H . Bock, Angew. Chem. 77, 469 (1965); Angew. Chem. internat.
Edit. 4,457 (1965).
[16] E. Konrad and L.Pelkns, Ber. Dtsch. Chem. Ges. 59,135 (1926).
Angew. Chem. internat. Edit.
Vol. 10 (1971) / No. 6
[17] H . W . Roesky, 0 . Glemser, and D. Bormann, Angew. Chem. 76,
713 (1964); Angew. Chem. internat. Edit. 3, 701 (1964).
[IS] M . Regitz, Angew. Chem. 79,786 (1967); Angew. Chem. internat.
Edit. 6, 733 (1967).
[I91 N . Wiberg and W.-Ch. Joo, unpublished.
[20] U . Krynitz, F. Gerson, N . Wiberg, and M . Veith, Angew. Chem. 81,
745 (1969); Angew. Chem. internat. Edit. 8, 755 (1969).
[21] If structure (8) occurred in equilibrium with ( 5 ) , (61, and (7),
it should be possible to intercept BSD e . g . with triphenylphosphane.
However, no reaction of BSD with this phosphane has been observed.
[22] K. Kimura, K . Katada, and J . H. Bauer, J. Amer. Chem. SOC.88,
416(1966).
[23] H . Kessler, Angew. Chem. 82,237 (1970); Angew. Chem. internat.
Edit. 9, 219 (1970).
[24] E. A . V . Ebsworth: Volatile Silicon Compounds. Pergamon Press,
Oxford 1963.
[25] The two K and n molecular orbitals formally result from a linear
combination of p, atomic orbitals and sp'-hybridized orbitals
respectively of the azo nitrogen atoms.
[26] M . B. Robin, R. R. Hart, and N . A . Kuebler, J. Amer. Chem. SOC.
89, 1564 (1967).
[27] E. Haselbach, J . A . Hashmall, E. Heilbronner, and V. Hornung,
Angew. Chem. 81, 897 (1969); Angew. Chem. internat. Edit. 8, 878
(1969).
[28] N. Wiberg, M . Veith, and H . Bachhuber, unpublished.
[29] a) The position of the highest occupied orbital of X-N=N-X
(X = CR,, SIR,) was established from the ionization energy (Table 1);
b) the assignment of the K+K* transition is still uncertain.
[30] H. Seidl, H . Bock, N . Wiberg, and M . Veith, Angew. Chem. 82,
42 (1970); Angew. Chem. internat. Edit. 9, 69 (1970).
[31] H . Schmidbaur and W . Malisch, Chem. Ber. 103,3007 (1970).
[32] Houben-Weyl: Methoden der organischen Chemie 10/2, 790
(1967).
[33] Chem. Eng. News 46, No. 33, 39 (1968).
[34] N . Wiberg and G. Schwenk, unpublished.
[35] N . Wiberg and W . Uhlenbrock, unpublished.
[36] N . Wiberg and W . Uhlenbrock, Angew. Chem. 82, 47 (1970);
Angew. Chem. internat. Edit. 9, 70 (1970).
[37] R.B. Woodward and R. Hofmann, Angew. Chem. 81,797 (1969);
Angew. Chem. internat. Edit. 8, 781 (1969).
[38] The thermolysis products of BSD always also contain some
bis(trimethylsily1) oxide, which is evidently formed by reaction of
BSD with the glass wall of the vessel.
[39] T.L. Cairns and B. C. McKusick, Angew. Chem. 73, 520 (1961).
[40] N . Wiberg,Angew. Chem. 80,809 (1968); Angew. Chem. internat.
Edit. 7, 766 (1968).
[41] N . Wiberg and W.-Ch. Joo, Z. Naturforsch., in press.
Angew. Chem. internat. Edit. 1 Vol. I0 (1971) 1 N o . 6
[42] Dimerization with formation of tetrazene, 2 BSD- +(R3Si)4Nz-,
is thought to be unlikely because of the accumulation of charge (cf.
the ion 0;). The tetrazane dianion could, however, be formed as a
transition state in the decomposition of Li[BSD] into nitrogen and (14).
1431 N . Wiberg, G. Schwenk, and W . Uhlenbrock, to be published.
[44] G. Briegleb: Elektronen-Donator-Acceptor-Komplexe, Springer,
Berlin 1961.
[44a] The Lewis base BSD acts in a way as an acid-base system, unlike
[E. Fahr and H. Lind, Anthe Lewis acid X-CO-N=N-CO-X
gew. Chem. 78,376 (1966); Angew. Chem. internat. Edit. 5,372 (1966)l.
[45] Since trimethylsilane cannot be detected as a product of the
thermolysis of BSD in toluene (Section 4.1), this reaction does not
proceed via R,SiH (cf. (14)).
[46] N. Wiberg and G . Schwenk, to be published.
[47] Ozone, which is homologous with SO,, reacts with BSD at -78°C
to form nitrogen, R,SiOH, and (R,Si),O. It is possible that, in analogy
with reaction (36), the compound R,Si-0-0-0-SiR,
is formed
first, and this would then decompose into (R,Si),O and singlet oxygen,
which itself reacts with BSD.
[48] W . I . Patnode and F. C. Schmidt, J. Amer. Chem. SOC.67, 2272
(1945); L. H . Sommer, E. W . Pietrusza, G . T. Kerr, and F. C . Whitemore,
ibid. 68, 156 (1946).
[49] N . Wiberg and H . J . Pracht, to be published.
[SO] The rate of the reaction of BSD with p-toluenesulfonyl azide is
lower than the rate of the BSD preparation reaction (2).
[SII N. Wiberg and B. Neruda, Chem. Ber. 99, 740 (1966).
1521 N . Wiberg and W.-Ch. Joo, J. Organometal. Chem. 22,349 (1970).
[53] E. W . Abel, R. A. N . McLean, and I . H . Sabherwal, J. Chem. Soc.
A 2371 (1968).
1541 A. Hughes and C . Sriuanauit, J. Heterocyclic Chem. 7, 1 (1970);
A. Schmidpeter and W . Zeiss, Angew. Chem. 83, 397 (1971); Angew.
Chem. internat. Edit. 10, 396 (1971).
1551 Under the hydrolysis conditions (2O-4o0C),only small quantities
of (11) and vanishingly small quantities of (12) are formed by
thermolysis of BSD. Most of the ( 1 1 ) and (12) that are produced in
large quantities during the hydrolysis must therefore be formed in
some other way.
[56] (9) and ( l o ) are also hydrolyzed to hydrazine and ammonium
azide by aqueous acids [36].
[57] Accordingly, hydrolysis intermediates such as (38) or (40) could
decompose into hydrazoic acid and silylamines (12) and (41), which
hydrolyze further to form ammonium azide.
[58] The hydrazine (37) is formed in large quantities in the protolysis
of BSD with a little ammonia.
[59] According to literature data [ll], the thermolysis of diimine
should lead only to nitrogen and hydrazine, so that the formation of
ammonium azide is apparently due not to the decomposition of diimine,
but necessarily to the hydrolysis of the silyltetrazenes (38) and (40).
387
Документ
Категория
Без категории
Просмотров
2
Размер файла
1 252 Кб
Теги
preparation, trimethylsilyl, structure, reactivity, bis, diimine
1/--страниц
Пожаловаться на содержимое документа