Mechanistic studies of a linear trisazoalkane, a new azimine, and a bicyclic triaziridine. Azoalkane homolysis into seven fragments

Article abstract of DOI:10.1021/jo991439g

An aliphatic azo compound containing three azo groups (1) has been prepared by IF₅ oxidation of β-azoamine 3. The thermolysis kinetics of this vicinal trisazoalkane were investigated above 155 °C, leading to a rate constant only 5.5 times faster than that of the simple model, azo-tert- butane. Because thermolysis to form seven stable products proceeds stepwise, the rate is hardly affected by the high exothermicity of the overall reaction (-93.4 kcal/mol). Oxidation of amine 3 also afforded a cyclic azimine 5 that underwent photolysis to yield a highly strained triaziridine 9 plus an unusual triazane 10, whose structures were elucidated by detailed NMR studies. On standing at ambient temperature, 9 reverted to 5 with a half-life of about an hour.

Full text of DOI:10.1021/jo991439g

1016
J . Org. Chem. 2000, 65, 1016-1021
Mech a n istic Stu d ies of a Lin ea r Tr isa zoa lk a n e, a New Azim in e,
a n d a Bicyclic Tr ia zir id in e. Azoa lk a n e Hom olysis in to Seven
F r a gm en ts
Paul S. Engel,* Li Pan, Kenton H. Whitmire, and Ilse Guzman-J imenez
Department of Chemistry, Rice University, P.O. Box 1892, Houston, Texas 77251
M. Robert Willcott
University of Texas Medical Branch at Galveston, 301 University Boulevard,
Galveston, Texas 77555-0793
William B. Smith
Department of Chemistry, Texas Christian University, Fort Worth, Texas 76129
Received September 13, 1999
An aliphatic azo compound containing three azo groups (1) has been prepared by IF5 oxidation of
â-azoamine 3. The thermolysis kinetics of this vicinal trisazoalkane were investigated above 155
°C, leading to a rate constant only 5.5 times faster than that of the simple model, azo-tert-butane.
Because thermolysis to form seven stable products proceeds stepwise, the rate is hardly affected
by the high exothermicity of the overall reaction (-93.4 kcal/mol). Oxidation of amine 3 also afforded
a cyclic azimine 5 that underwent photolysis to yield a highly strained triaziridine 9 plus an unusual
triazane 10, whose structures were elucidated by detailed NMR studies. On standing at ambient
temperature, 9 reverted to 5 with a half-life of about an hour.
Ta ble 1. Th er m olysis Activa tion P a r a m eter s of Selected
Azoa lk a n es
While a number of bisazoalkanes appear in the litera-
ture,1 very few compounds containing three aliphatic azo
moieties have been reported.2 The high stability of vicinal
bisazoalkane 61,3 encouraged us to examine the related
∆Hq,
∆Gq,
a
compound solvent
kcal/mol
∆Sq, eu
16.3
kcal/mola krel
ATBb
Ph2O-iqc 42.1
34.7
33.9
33.2
1.0
6d
1
toluene
decane
35.2 ( 0.3
35.8 ( 0.2
2.8 ( 0.6
5.8 ( 1.5
2.4
5.5
a
d
At 180.0 °C. b Data from ref 10. c iq is isoquinoline. Data from
ref 1.
unreacted. Treatment of 3 with sulfuryl chloride6 afforded
sulfamide 4, which was oxidized by the nonaqueous
hypochlorite-base method7 to yield azimine 5. At -78 °C,
trisazoalkane 1, whose highly exothermic decomposition
could produce seven fragments. In the course of synthe-
sizing 1, we encountered a cyclic azimine 5 that closed
photochemically to a highly strained triaziridine.
8,9
addition of IF5 to 3 afforded 1 as a stable, light yellow
solid in 32% yield but addition of 3 to IF5 at higher
temperatures led only to 5.
Taking advantage of Ciganek’s valuable transforma-
tion4 of nitriles to tert-alkylamines, we converted the
known cyanoazoalkane 25 to â-azoamine 3. Prolonged
reaction times led to formation of acetone tert-butyl
hydrazone, which presumably arose by scission4 of the
iminyl anion from addition of one methyl group to 2.
Since this hydrazone was not readily separable from 3,
the reaction conditions were carefully controlled to
minimize its yield, even though a good deal of 2 remained
Th er m olysis of 1. A solution containing 0.124 mmol
of 1 in 1.5 mL of decane was heated at 170 °C for 10 h,
yielding 97.5% of the nitrogen expected from loss of 3
equiv of N2. Thermolysis kinetics of 1 in degassed decane
were carried out in sealed 1 cm Pyrex cuvettes, which
were thermolyzed in a regulated oil bath and monitored
by UV spectroscopy. Seven temperatures between 155
and 182 °C were employed, giving clean first-order
kinetics and a linear Eyring plot (F ) 0.999). The
activation parameters are compared with those of 6 and
azo-tert-butane (ATB) in Table 1.
(1) Engel, P. S.; Wang, C.; Chen, Y.; Ru¨chardt, C.; Beckhaus, H.-D.
J . Am. Chem. Soc. 1993, 115, 65-74.
(2) Sheppard, C. S.; MacLeay, R. E. Symmetrical Trisazo Sequential
Free Radical Initiators Having Two Different Azo Functions. U. S.
Patent 3,868,359, Feb. 25, 1975.
(3) Engel, P. S.; Chen, Y.; Wang, C. J . Am. Chem. Soc. 1991, 113,
4355-4356.
(4) Ciganek, E. J . Org. Chem. 1992, 57, 4521-4527.
(5) MacLeay, R. E.; Sheppard, C. S. Unsymmetrical Azo and
Hydrazo Compounds. Canadian Patent 924299, April 10, 1973.
To probe the thermolysis mechanism of 1, a solution
of 10 mg of 1 and 12 mg of thiophenol in 0.5 mL of
(6) Stowell, J . C. J . Org. Chem. 1967, 32, 2360-2362.
(7) Timberlake, J . W.; Hodges, M. L.; Betterton, K. Synthesis 1972,
632-634.
10.1021/jo991439g CCC: $19.00 © 2000 American Chemical Society
Published on Web 02/02/2000

Azoalkane Homolysis into Seven Fragments
J . Org. Chem., Vol. 65, No. 4, 2000 1017
benzene was thermolyzed for various times at 153.56 °C.
GC/MS analysis revealed both 7 and 8, and the structure
of 7 was confirmed by comparison with an authentic
sample.
kinetics of the regeneration of 5 were determined by
subtracting the immediate postirradiation spectrum from
each spectrum obtained as the solution aged. This
correction led to an array of azimine spectra over time
with an invariant λmax of 288 nm, which reappeared with
a rate constant of 3.84 × 10-4 s-1 at 23.4 °C.
The photochemical ring closure of 5 could also be
monitored by 1H NMR. Irradiation (313 nm) of 5 in CDCl3
below 0 °C decreased its two singlets at 1.02 and 1.47
ppm and caused the appearance of two sets of three
singlets each, with the peaks of each set in the ratio 3:2:
2. The larger of these sets (0.93, 1.04, and 1.24 ppm) was
assigned to the anti triaziridine 9 while the smaller one
(1.01, 1.83, and 1.87 ppm) was assigned to triazane 10.
Both 9 and 10 proved to be thermally labile, with 9
disappearing much faster than 10. Monitoring the 0.93
ppm singlet of 9 in the NMR probe at a temperature of
24.4 °C led to a disappearance rate constant of 2.05 ×
10-4 s-1 (F ) 0.999). The sum of the NMR peak areas of
5 (t-Bu at 1.47 ppm) and 9 (t-Bu at 0.93 ppm) gave a
value that rose by less than 3% as >90% of 9 disappeared,
indicating that the only fate of 9 is reversion to 5. The
reappearance rate of 5 based on its 1.47 ppm singlet was
1.90 × 10-4 s-1, but this value is less reliable than the
disappearance rate of 9 because the infinity point was
estimated. The two-fold discrepancy between the UV and
NMR regeneration rate of 5 is attributed to the different
solvents used and possible temperature measurement
errors.
Th e Ch em istr y of Azim in e 5. The unexpected az-
imine 5 turned up in oxidation both of the sulfamide 4
and of the azoamine 3 at warmer temperatures. This
compound was a stable liquid whose high-resolution EI
mass spectrum gave the formula C10H21N3 and whose UV
showed a band at 288 nm (ꢀ ) 2500), similar to that seen
in other azimines.11-13 The 1H NMR spectrum of 5
exhibited only two singlets at 1.02 and 1.47 ppm in the
ratio of 4:3. In the 15N-1H HETCOR spectrum, the larger
singlet at 1.02 ppm correlated with the 15N signal at 294.9
1
ppm relative to NH3 while the smaller H singlet at 1.47
ppm correlated with the 15N signal at 326.5 ppm, which
must therefore be the central azimine nitrogen. An
unequivocal proof of structure followed from single-
crystal X-ray diffraction studies of the stable picrate salt
of 5 (mp 129.5 °C) and of the HI salt, which was isolated
as crystals from IF5 oxidation of 3.
Azimines are known to proceed photochemically to
triaziridines.14 We irradiated 5 in hexane at 313 nm and
0 °C in an attempt to form anti triaziridine 9 and/or its
syn isomer. The characteristic 288 nm band of 5 dimin-
ished rapidly, leaving a weaker 270 nm band. As the
solution stood in a thermostated cell holder at 23.4 °C,
the 288 nm band gradually reappeared. Since there was
no isosbestic point, the 270 nm band is not due to a
triaziridine but instead must represent another photo-
chemical product, probably triazane 10 (see below). The
The other irradiation product of 5 exhibited chemical
behavior consistent with its structural assignment as 10.
After 36 min irradiation of 5, the photolyzate was quickly
filtered through Celite and silica gel to remove cloudiness.
This treatment destroyed 10 and converted it to acetone,
1
which was obvious from its 2.09 ppm H singlet and its
205.74 ppm 13C peak. Continued irradiation at 0 °C
caused the buildup of both 9 and 10. The thermolysis of
10 was monitored by its 1.83 ppm NMR singlet, leading
to a rate constant of 1.03 × 10-5 s-1. After several days
at room temperature, the NMR spectrum of the product
mixture from 5 exhibited many peaks whose origin could
have been 10.
The assignment of structure to the two photoisomers,
while seemingly easy, required a careful analysis of the
1H, 13C, and 15N correlated NMR spectra. First we note
that the spectra of both 9 and 10 exhibit the same
symmetry. There are two different methyl resonances (six
(8) Stevens, T. E. J . Org. Chem. 1961, 26, 2531-2533.
(9) Timberlake, J . W.; Pan, D.; Murray, J .; J ursic, B. S.; Chen, T. J .
Org. Chem. 1995, 60, 5295-5298.
(10) Engel, P. S. Chem. Rev. 1980, 80, 99-150.
(11) Hilpert, H.; Dreiding, A. S. Helv. Chim. Acta 1988, 71, 277-
287.
(13) Smith, R. H.; Pruski, B.; Day, C. S.; Ptaltzgraff, T. D.; Michejda,
C. J . Tetrahedron Lett. 1992, 33, 4683-4686.
(14) Hilpert, H.; Hoesch, L.; Dreiding, A. S. Helv. Chim. Acta 1985,
68, 1691-1697.
(12) Storr, R. C. Azimines, Azoxy Compounds, and Nitro Compounds;
Padwa, A., Ed.; Wiley: New York, 1984; Vol. 2, pp 169-196.

1018 J . Org. Chem., Vol. 65, No. 4, 2000
Sch em e 1. Th er m olysis of 1 w ith P h SH
Engel et al.
hydrogens total in each), one tert-butyl proton resonance,
13C peaks for the methyl and tert-butyl groups, two
quaternary methyl signals in the intensity ratio 2:1, and
two nitrogen resonances in the ratio of 2:1. Triaziridine
9 exhibited 15N chemical shifts similar to those of related
triaziridines,15 and its configuration was assigned as anti
on the basis of the absence of any NOE between the tert-
butyl and ring methyl groups. The signals in each isomer,
9 and 10, were correlated by direct and long-range 2D
heteronuclear NMR.16 In 10, the nitrogen signal of
intensity 2 at 179.7 ppm and the quaternary carbon of
intensity 2 at 165.57 ppm are in the chemical shift range
expected for the NdC group. The details of the correlated
spectra verify the structures assigned to 9 and 10, which
are fully rationalized elsewhere.16 An online search of
Beilstein uncovered only two compounds (11a ,b), whose
structure resembles 10.17
Breaking bond a in Scheme 1 affords a tert-butyl radical
plus â-azo radical 13, which fragments to 12, while initial
cleavage of bond b leads directly to two molecules of 12.
These â-azo radicals are already known to undergo rapid
fragmentation to tetramethylethylene (TME), nitrogen,
and a tert-butyl radical.3
The overall thermolysis of 1 into seven fragments is
considerably exothermic. Whereas fragmentation of ATB
is endothermic by 33.3 kcal/mol (eq 1), 6 is nearly
thermoneutral (eq 2) and 1 releases 33.1 kcal/mol (eq 3).
The major fate of tert-butyl radicals is disproportionation
to isobutane plus isobutene, a reaction that liberates 60.3
kcal/mol. Combining this figure with ∆Hr of reactions 1
and 3, we find that thermolysis of 1 to seven stable
products liberates 93.4 kcal/mol, more than three times
as much energy as thermolysis of azo-tert-butane to yield
isobutane plus isobutene (∆Hr ) -27.0 kcal/mol.) Ex-
trapolation of eqs 1-3 indicates that each additional
t-Bu-NdN-CMe2 group renders homolysis more exother-
mic by ∼33 kcal/mol. It is important to note that the
stepwise nature of these thermolyses masks most of the
effect of exothermicity on the activation enthalpy.
Discu ssion
Thermolysis studies of 1 demonstrated that despite the
presence of three azo groups, this compound is very stable
at ambient temperature. On a statistical basis, 6 should
decompose twice as fast as ATB, and 1 should proceed
at three times the rate of ATB. The fact that the
decomposition rates of 6 and 1 were faster than these
figures is attributed to a higher ground-state energy
caused by greater steric crowding.18,19
The formation of azoalkanes 7 and 8 in the thermolysis
of 1 with thiophenol shows that both radicals 12 and 13
are intermediates; that is, homolysis proceeds stepwise.
(15) Hilpert, H.; Hollenstein, R. Helv. Chim. Acta 1986, 69, 136-
140.
(16) Luxon, B. A.; Willcott, M. R.; Ezell, E. L. Submitted for
publication.
(17) Blackstock, D. J .; Happer, D. A. R. Chem. Commun. 1968, 63.
(18) Garner, A. W.; Timberlake, J . W.; Engel, P. S.; Melaugh, R. A.
J . Am. Chem. Soc. 1975, 97, 7377-7379.
(19) Ruchardt, C. Top. Curr. Chem. 1980, 88, 1-32.

Azoalkane Homolysis into Seven Fragments
J . Org. Chem., Vol. 65, No. 4, 2000 1019
Sch em e 2. P h otoch em ica l a n d Th er m a l
Rea ction s of 5 a n d 9
The only observed thermal reaction of 9 is simply
reversion to 5, a process found in other triaziridines.11
This ring opening is much more facile than the conversion
of bicyclo[2.1.0]pentane to cyclopentene, which proceeds
through an unstabilized 1,3-biradical.
DFT calculations on the opening of 9 to 5 indicated a
reaction exothermicity of 45 kcal/mol and a low activation
energy. A large number of attempts to locate the TS for
the ring-opening process led to many structures meeting
the requirement of a single imaginary frequency. How-
ever, these invariably reflected pseudorotations of the
five-membered ring or one or more of the seven methyl
groups. The PE surface for the ring opening was very
flat with each of these TS’s appearing as pimples on the
surface. Subsequently, to approximate the TS for the
bond-breaking process, a series of 30 structures was
generated by a relaxed potential energy in which the
breaking N-N bond was extended from 1.5 to 2.15 Å in
small increments (ca. 150 h CPU time). A possible
maximum in the energy was found at or near 1.780 Å,
corresponding to a structure whose central nitrogen had
progressed only slightly toward planar. This proved not
to be a true TS, but the energy was assumed to be close
to that of the desired structure. The approximate activa-
tion enthalpy was 11.0 kcal/mol, which is lower than the
22.4 kcal/mol calculated from the disappearance rate of
9 at 24.5 °C with an assumed activation entropy of zero.
In summary, trisazoalkane 1 undergoes exothermic
thermolysis to seven fragments. Despite its ∼65 kcal/mol
greater thermodynamic driving force for homolysis than
that of the simple model, azo-tert-butane, its rate is not
even six times faster than that of ATB. A byproduct of
the synthesis, azimine 5, rearranged to 9 on irradiation,
and this strained bicyclic triaziridine reopened to 5 with
a half-life of about an hour at ambient temperature.
Another irradiation product of 5 was found to be the
triazane 10, a member of a rare class of compounds.
The reaction enthalpies in eqs 1-3 were determined
from the known ∆Hf of ATB10 and t-Bu‚,20 the experi-
mental value for tetramethylethylene,21 and the calcu-
lated ∆Hf of 1 and 6. To obtain the latter, we assume
that the heat of hydrogenation of each quaternary C-C
bond of 1 and 6 equals that of 14, which is -8.5 kcal/mol
(cf. eq 4).1 ∆Hf of t-Bu-NdN-i-Pr is calculated as the
average of the values for ATB and azoisopropane.10
A possible mechanism for the formation of azimine 5
during the synthesis of 1 is halogenation of the amino
group of 3 or of the sulfamide nitrogen of 4 followed by
displacement of the halogen by the azo group. However,
since no mechanistic studies have been carried out, we
cannot rule out 9 as an intermediate in the formation of
5.22 The new azimine afforded both a picrate and a
hydrogen iodide salt, selected bond lengths of which are
shown below. Those of the iodide are listed above those
of the picrate, which are in parentheses. The N1-N2
bond is shorter than the N2-N3 bond, indicating pro-
tonation of N3 and a double N1-N2 bond. The N3-X
bond length is greater in the iodide (X ) I) than in the
picrate (X ) O), presumably an effect of atomic size. On
account of the net positive charge on the azimine moiety,
Exp er im en ta l Section
1
Gen er a l. H NMR spectra were run at 250 MHz in CDCl3
unless otherwise specified. Other general procedures were as
1
the H NMR chemical shifts of the picrate and iodide lie
described earlier.25
downfield from those in free 5. The observation that the
spectrum of the salts consists only of two singlets implies
a plane of symmetry due to rapid exchange of the proton
from N1 to N3.
2-(ter t-Bu tylh yd r a zo)-2-cya n op r op a n e5 was prepared by
adding acetone (4.2 mL, 0.057 mol) dropwise to a stirred
solution of t-BuNHNH3Cl (7.15 g, 0.057 mol) and NaCN (2.81
g, 0.057 mol) in 20 mL of deionized water. The flask was
stoppered, and the mixture was stirred overnight, whereupon
the product separated an oil. The aqueous layer was extracted
with CH2Cl2, and the combined organic layer was dried over
Na2SO4. The solvent was removed by rotary evaporation,
leaving a yellow oil (yield > 95%). 1H NMR δ 1.03 (s, 9H), 1.41
(s, 6H).
A plausible mechanism for the reactions of azimine 5
is shown in Scheme 2. Irradiation of 5 affords 9, a
reaction that can be classified as an allowed disrotatory
process.23 Triazane 10 might arise directly by photolysis
of 5 or by thermolysis of the exceedingly strained and
undetected syn isomer of 9. If this syn triaziridine is an
intermediate, its rapid thermal ring opening to 10 is
analogous to a known minor reaction of bicyclo[2.1.0]-
pentane, conversion to 1,4-pentadiene.24 Attempts to
model the syn isomer calculationally led to a spontaneous
tert-butyl flip to the thermodynamically more stable 9.
2-ter t-Bu tyla zo-2-cya n op r op a n e (2)5 was made by adding
132 mL of commercial bleach (5.25% NaOCl) over a period of
1 h to 3.6 g (0.023 mol) of ice cold 2-(tert-butylhydrazo)-2-
cyanopropane with stirring. The ice bath was removed, and
the reaction mixture was stirred for an additional 3 h. After
collection of the upper layer consisting of product 2, the
aqueous layer was extracted with ether. The combined organic
layer was washed with 1% HCl, followed by two portions of
saturated NaHCO3 and one of deionized water. The organic
layer was dried over Na2SO4, and the ether was rotary evap-
orated, leaving a pale yellow oil (yield > 90%). 1H NMR δ 1.22
(s, 9H), 1.61 (s, 6H). 13C NMR (CDCl3) 25.03, 26.41, 67.36,
(20) Berkowitz, J .; Ellison, G. B.; Gutman, D. J . Phys. Chem. 1994,
98, 8, 2744-2765.
(21) Pedley, J . B.; Naylor, R. D.; Kirby, S. P. Thermochemical Data
of Organic Compounds, 2nd ed.; Chapman and Hall: London, 1986.
(22) Marterer, W.; Klinger, O.; Thiergardt, R.; Beckmann, E.; Fritz,
H.; Prinzbach, H. Chem. Ber. 1991, 124, 621-633.
(23) Woodward, R. B.; Hoffmann, R. The Conservation of Orbital
Symmetry; Verlag Chemie: Weinheim, Germany, 1970; p 57.
(24) Steel, C.; Zand, R.; Hurwitz, P.; Cohen, S. G. J . Am. Chem. Soc.
1964, 86, 679-687.
67.76, 120.5. IR 2991, 2240, 2220, 1463, 1364, 1224 cm-1
.
(25) Billera, C. B.; Dunn, T. B.; Barry, D. A.; Engel, P. S. J . Org.
Chem. 1998, 63, 9763-9768.

1020 J . Org. Chem., Vol. 65, No. 4, 2000
Engel et al.
2-Am in o-2,3-d im eth yl-3-(ter t-bu tyla zo)bu ta n e (3). The
method was modified from Ciganek4 and Imamoto.26 Cerium
chloride heptahydrate27 (10 g) was quickly ground to a fine
powder in a warm, oven-dried mortar. The powdered CeCl3
was transferred into a 250 mL flask containing a magnetic
stirring bar and a three-way stopcock. The flask was evacuated
to 0.2 mmHg and was gradually heated in a 100 °C oil bath
without magnetic stirring for 2.0 h. Then the bath temperature
was raised to 150 °C, the CeCl3 was dried with magnetic
stirring for another 1 h, and the sublimed solids were returned
into the flask by frequent tapping. The vacuum was replaced
by argon gas while the flask was still hot. The flask was first
cooled in the air and then in an ice bath; then cool, freshly
distilled, dry THF (100 mL) was added. After the suspension
was stirred at room temperature overnight, the flask was
cooled to -78 °C and 1 equiv of CH3Li (1.6 M in Et2O) was
added. The mixture was stirred for 1-1.5 h at -78 °C; then
the bath temperature was raised to -65 °C. 2-tert-Butylazo-
2-cyanopropane 2 (0.25 equiv) was added, and stirring was
continued for 1.5 h at -65 °C. Concentrated NH4OH (30 mL)
was added, and the mixture was allowed to warm to room
temperature. The cerium slurry was filtered and washed with
ether. The solvent was removed by rotary evaporation, and 5
mL of 5% aqueous HCl was added with stirring to the crude
product. The mixture was extracted with ether to remove
unreacted 2. Concentrated aqueous NaOH was added very
slowly to the remaining aqueous solution in an ice bath until
the pH rose to 9. The azoamine was extracted with ether and
dried over K2CO3. Removal of the solvent by rotary evaporation
left 3 as a yellow oil in 30% yield. 1H NMR δ 1.04 (s, 6H), 1.13
(s, 6H), 1.17 (s, 9H). 13C NMR (CDCl3) 20.39, 26.19, 26.77,
54.69, 66.91, 73.03. IR 3369, 3286, 2967, 1698, 1462, 1358,
1205, 804 cm-1. HRMS CI+ (M + H) 186.197 03, calcd
186.197 02. UV λmax ) 372 nm.
2,3-Dim et h yl-3-(ter t-b u t yla zo)b u t -2-yl su lfa m id e (4)
was made according to Stowell’s method.6 A 1 g portion of
2-amino-2,3-dimethyl-3-(tert-butylazo)butane 3 in 2 mL of dry
pentane was cooled in an ice bath under N2. SO2Cl2 (109 µL)
in 0.2 mL of pentane was added over a period of 30 min. After
removal of the ice bath, the mixture was stirred for 2 h.
Deionized water (0.5 mL) and ether (2 mL) were added slowly,
and mixture was stirred for 5 min. The organic layer was
washed with 5% aqueous HCl and brine. Removal of the
solvent yielded a light yellow solid, which was recrystallized
from pentane. Yield 48%. 1H NMR δ 1.06 (s, 6H), 1.18 (s, 9H),
1.46 (s, 6H) 13C NMR (CDCl3) 20.37, 22.61, 26.65, 61.85, 67.60,
73.44. HRMS CI+ (M + 1H) 433.331 62, calcd 433.332 47.
Azim in e 5. An 8 mg portion of a 50% dispersion of NaH in
mineral oil (167 µmol) was rinsed with pentane and stirred
with 0.5 mL of dry pentane. Sulfamide 4 (25 mg, 57.8 µmol)
in 0.2 mL of dry pentane was added slowly to the ice cold NaH
suspension. The mixture was stirred at room temperature for
4 h and then recooled to 0 °C. tert-Butyl hypochlorite (15 µL,
126 µmol, in 25 µL of pentane) was added dropwise, the
temperature was allowed to rise to ambient, and the mixture
was stirred overnight. Water was added very slowly to destroy
the excess NaH. The pentane layer was separated, dried over
Na2SO4, and carefully concentrated by microdistillation. Pure
5 was obtained by silica gel chromatography, eluting with 1:4
Et2O:pentane. Yield 60%.
night. The yellow HI salt of 5 (160 mg) crystallized out of the
oil. After filtration, the mother liquor was added to 10 mL of
CH2Cl2 and washed with 3 × 5 mL brine. The solvent was
removed and the residue subjected to silica gel chromatogra-
1
phy, yielding 760 mg (85%) of liquid azimine 5. H NMR (400
MHz, CDCl3) δ 1.02 (s, 12H), 1.47 (s, 9H). 13C NMR (CDCl3)
1
19.14, 27.98, 64.41, 69.18. These H and 13C δ’s differ slightly
from those in the Supporting Information, which were obtained
at 250 MHz. 15N NMR (CDCl3 solvent, δ relative to NH3) 294.9,
326.5. IR 2973, 2934, 1514, 1448, 1415, 1363, 1226, 1213, 1167
cm-1. UV λmax ) 288 nm, ꢀ ) 2500. HRMS EI+ 183.1732, calcd
183.1732. HI salt: 1H NMR δ 1.41 (s, 12H), 1.85 (s, 9H).
P icr a te of 5. In a 3 mL conical vial, 0.3 mL of a saturated
solution of picric acid in 95% EtOH was added to 15 mg of
azimine 5 in 0.3 mL of 95% EtOH. The vial was equipped with
a condenser, and the mixture was refluxed for 1 min. The
solution was cooled slowly to room temperature. Needlelike
yellow crystals were collected by vacuum filtration. The
crystals were further purified by recrystallization from 95%
EtOH. Mp 129.5 °C. 1H NMR 1.31 (12H), 1.75 (9H), 8.91 (2H).
Tr isa zoa lk a n e 1 was made according to a modified proce-
dure of Timberlake.9 A 100 mL flask charged with 1.6 g of 3,
9.2 mL of dry pyridine (freshly distilled from CaH2), and 30
mL of dry CH2Cl2 was quickly connected to a 20 mL funnel
equipped with a septum and an argon balloon. An IF5 solution
(19 mL of 3.5 g of IF5 in 25 mL of CH2Cl2) was quickly syringed
into the funnel. The IF5 solution was added over 4 h at -78
°C, and the resulting solution was stirred for 5.5 h at -78 °C.
The bath temperature was raised slowly to -50 °C, 2 mL of
ethanol was added dropwise, and the bath temperature was
raised to -10 °C over 15 min. Water (5 mL) was added
dropwise, and the mixture was stirred for 10 min; then, the
mixture was poured into ice water. The organic layer was
washed successively with 5% HCl, 5% NaHSO3, and brine;
then it was dried over Na2SO4. Following removal of the
solvent, silica gel chromatography with 5% ether in pentane
afforded 1 as a yellow solid. Recrystallization from MeOH and
1
Et2O (9:1) gave light yellow flakes in 32% yield. Mp 75 °C. H
NMR (C6D6) δ 1.21 (18H), 1.29 (12H), 1.34 (12H). 1H NMR
(CDCl3) δ 1.106 (12H), 1.133 (12H), 1.158 (18H). 13C NMR
(CDCl3) 20.76, 20.87, 26.80, 66.69, 72.67, 73.67. HRMS CI+
(M + 1H) 367.35454, calcd 367.35492. UV (λmax ) 374 nm, ꢀ )
55). 150 mg of 5 was obtained as a byproduct in this reaction.
Th er m olysis P r od u cts of 1. The 1H NMR spectrum of
thermolyzed samples of 1 in C6D6 exhibited major peaks for
isobutane (0.863 d, 10H, J ) 6.6 Hz; 1.637 dectet, 1H, J ) 6.6
Hz), isobutene (1.596 t, 6H, J ) 1.15 Hz; 4.750 sept, 2H, J )
1.14 Hz), and tetramethylethylene (1.618 s), as judged by
comparison with authentic samples. An unidentified doublet
appeared at 0.964 ppm, J ) 6.8 Hz. In a separate experiment,
a benzene solution containing 0.055 M 1 and 0.23 M thiophenol
was placed into three tubes, which were degassed and sealed.
After heating for a known time, the tubes were opened and
analyzed by GC using an internal standard. The times (min)
and concentrations (M) of 7 and 8 (M) were 262, 0.0020, 0.0010;
501, 0.0020, 0.00067; 1366, 0.0013, 0.000065.
Nitr ogen Yield of 1. A solution of 15.1 mg of 1 in 1.5 mL
of decane was subjected to four freeze-thaw degas cycles at
77 K below 10-4 mmHg. The tube was sealed under vacuum
and was heated at 170 °C for 10 h. After the tube was broken
under vacuum, the evolved gases were collected using a To¨pler
pump with a liquid nitrogen trap to retain any condensable
gases. The nitrogen, which was quantified in a gas buret,
corresponded to a yield of 97.5%.
Th er m olysis Kin etics of 1. The disappearance rate of 1
was monitored by UV using a HP-8452 diode array spectrom-
eter. The temperatures (°C) and rate constants (s-1) are as
follows: 155.49, 8.81 × 10-5; 157.42, 1.04 × 10-4; 161.97, 1.74
× 10-4; 169.85, 3.33 × 10-4; 173.92, 4.90 × 10-4; 178.62, 8.07
× 10-4; 182.45, 1.16 × 10-3. These values were used in the
Eyring equation to obtain the activation parameters in Table
1.
In an alternate procedure, 5 was made by adding 1 mL of
dry pyridine and 1.35 g of IF5 to 8 mL of dry CH2Cl2 under
argon. The solution was stirred for 40 min at -78 °C.
Azoamine 3 (1.0 g) in 3 mL of CH2Cl2 was added at -40 °C,
and the mixture was stirred for 20 min. The temperature was
allowed to rise to 0 °C, and stirring was continued for 2 h.
Water (3 mL) was added dropwise, and the mixture was stirred
until all of the precipitate had dissolved. The organic layer
was washed with 5% HCl and then 5% NaHSO3. CH2Cl2 was
evaporated, and the oily product was stored at -5 °C over-
(26) Imamoto, T. Organocerium Reagents; Imamoto, T., Ed.; Per-
gamon Press: Elmsford, NY, 1991; Vol. 1, pp 231-250.
(27) Dimitrov, V.; Kostova, K.; Genov, M. Tetrahedron Lett. 1996,
37, 6787-6790.
Ir r a d ia tion P r od u cts of 5. A solution of 5 cooled below 0
°C was irradiated with 313 nm light from a 500 W high-
pressure mercury lamp through a potassium chromate filter

Azoalkane Homolysis into Seven Fragments
J . Org. Chem., Vol. 65, No. 4, 2000 1021
solution. The irradiation in pentane (10 mL) employed 50 mg
of 5. For the NMR experiment, a solution of 90 mg of 5 in 0.4
mL of CDCl3 was monitored by periodic spectra taken as
quickly as possible to avoid reversion of 9 to 5. Triaziridine 9
1H NMR (400 MHz, CDCl3) 0.93 (9H), 1.04 (6H), 1.24 (6H).
13C NMR 20.16, 21.80, 23.70, 54.96, 64.19. 15N NMR versus
NH3 148.6, 154.0. Triazane 10 1H NMR 1.01 (9H), 1.83 (6H),
1.87 (6H). 13C NMR 18.9, 24.3, 24.5, 60.44, 165.57. 15N NMR
versus NH3 179.7, 340.7.
minimized at the HF/3-21G* level. Such minimizations rou-
tinely required 4-6 h on a Dec-Alpha 500 MHz workstation.
The relaxed potential energy scan (30 data points) was also
carried out at this level. All structures were then reoptimized
at the Becke3LYP/6-31G* level following a suggestion from
Professor Paul v. R. Schleyer.29
Ack n ow led gm en t. We thank the National Science
Foundation and the Robert A. Welch Foundation for
support of this work.
Th eor etica l Ca lcu la tion s. All calculations were carried
out with Gaussian 98.28 The structures of 5 and 9 were initially
(28) Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.;. Cheeseman, J . R.; Zakrzewski, V. G.; Montgomery, J r.,
J . A.; Stratmann, R. E.; Burant, J . C.; Dapprich, S.; Millam, J . M.;
Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J .;
Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo,
C.; Clifford, S.; Ochterski, J .; Petersson, G. A.;. Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J . B.; Cioslowski, J .; Ortiz, J . V.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.;. Martin, R.
L.; Fox, D. J .; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara,
A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; J ohnson, B.; Chen,
W.; Wong, M. W.; Andres, J . L.; Gonzalez, C.; Head-Gordon, M.;
Replogle, E. S.; Pople, J . A. Gaussian 98, Revision A.4; Gaussian,
Inc.: Pittsburgh, PA, 1998.
Su p p or tin g In for m a tion Ava ila ble: A table of SPAR-
TAN quantum chemical results, X-ray crystallography data
for the picrate and hydrogen iodide salts of azimine 5, and
NMR spectra of 1 and 5. This material is available free of
charge via the Internet at http://pubs.acs.org.
J O991439G
(29) Private communication.