Boundary conditions on the partitioning of deaminatively generated benzyl cations

Article abstract of DOI:10.1021/jo001458m

Nitrogenous entity-separated ion pairs (NESIPs) containing benzyl cations, nitrogen gas, and pivalate anions were generated via thermal deamination of N-benzyl-N-nitrosopivalamide. Some decompositions were performed in methanolic solutions saturated with

Full text of DOI:10.1021/jo001458m

J . Org. Chem. 2001, 66, 2681-2685
2681
Bou n d a r y Con d ition s on th e P a r tition in g of Dea m in a tively
Gen er a ted Ben zyl Ca tion s
Ron W. Darbeau,* Derrill J . Heurtin,§ Luis M. Siso,† Rebecca S. Pease, Rebekah E. Gibble,
Davin E. Bridges, Daneisha L. Davis,‡ Delana G. Gbenekama,‡ and Nyla R. Darbeau
Department of Chemistry, McNeese State University, Lake Charles, Louisiana, 70609
rdarbeau@mail.mcneese.edu
Received October 10, 2000
Nitrogenous entity-separated ion pairs (NESIPs) containing benzyl cations, nitrogen gas, and
pivalate anions were generated via thermal deamination of N-benzyl-N-nitrosopivalamide. Some
decompositions were performed in methanolic solutions saturated with selected nucleophiles:
acetate, azide, or cyanide ions. Trace amounts of benzyl cyanide and tolunitriles were observed; no
corresponding products were detected in the acetate and azide cases. Other decompositions were
performed in the absence of traditional solvent but in the presence of the nucleophilic salts; again
only poor cyanide interception of the cation was observed. The poor showing of the nucleophilic
ions, when present, is discussed in the context of the lifetime of the cation, effective nucleophilicity,
and cage effects in deamination.
In tr od u ction
The ability of nucleophilic solvents such as benzene/
toluene mixtures,1a,b pyrrole,1c acetonitrile,1d styrene,1e
and acetone1f to intercept deaminatively generated car-
bocations as probes of reaction mechanisms1a-c and to
form novel and useful products in variable yields (7%1a,b
to 78%1c) has been recently explored. These carbocations,
formed via unimolecular1g,h thermolysis of N-alkyl-N-
nitrosoamides as part of a nitrogenous-entity separated
ion-pair (NESIP; Figure 1), are very reactive.1,2 Their
enhanced reactivity is due, in part, to the low activation
energy required for the loss of the nitrogenous molecule
from the alkyldiazonium precursor thus allowing the
cation to be formed with minimal solvent participation.1,2
Additionally, the temporary screening of the cation from
its counterion by the sheer physical presence of the
nitrogenous entity results in the maximal positive charge
at the electron-deficient center.1,2
F igu r e 1. Initial structure of a NESIP containing alkyl and
carboxylate ions.
nitrosonaphthamide2b (and its 2-butyl analogue2b) and
18O-studies of “intramolecular inversion”2b require the
finite existence of a carbocation. Additionally, (2) deami-
nation of bridgehead amines leads to solvent-derived
product (SDP),2c,e which is impossible if the diazonium
ion were the alkylating agent.
The alkyl cation is believed to be the active electrophile
because (1) predominant retention of configuration ob-
served in the ester from deamination of phenylethyl-N-
The absence of a free-radical pathway is indicated by
the nondetection of monodeuteriomethylbenzene or
hexachloroethane from the decomposition of N-benzyl-
N-nitrosobenzamide in 10:1 benzene/CDCl3 at 80 °C.2d
Since no D-abstraction from CDCl3 occurred at a mod-
estly high temperature in a fairly nonpolar medium, then
a free-radical pathway is apparently not competitive with
the cationic pathway. Further, the similarity among kT/
kB and isomer distribution values from nitrosoamide
decompositions in equimolar benzene-toluene at 80, 40,
and 25 °C1b,2d also indicate the absence of a competitive
radical pathway.
* To whom correspondence should be addressed.
§ Chemistry teacher, Alfred M. Barbe High School, Lake Charles,
LA, 70605.
† Undergraduate Chemistry Major at McNeese State University.
‡ American Chemical Society, Project SEED students.
(1) (a) White, E. H.; Darbeau, R. W.; Chen, Y.; Chen, D.; Chen, S.
J . Org. Chem. 1996, 61, 7986. (b) Darbeau, R. W.; White, E. H.; Song,
F.; Darbeau, N. R.; Chou, J . C. J . Org. Chem. 1999, 64(16), 5966. (c)
Darbeau, R. W.; White, E. H. J . Org. Chem. 1997, 62, 8091. (d)
Darbeau, R. W.; White, E. H.; Nunez, N. P.; Coit, B. C.; Daigle, M. A.
J . Org. Chem. 2000, 65, 1115. (e) Darbeau, R. W.; Delaney, M. S.;
Ramelow, U.; J ames, K. R. Org. Lett. 1999, 1(5), 761. (f) Song, F.;
Darbeau, R. W.; White, E. H. J . Org. Chem. 2000, 65, 1825. (g) Huisgen,
R.; Ruchardt, C. J ustus Liebigs. Ann. Chem. 1956, 601, 1. (h) Darbeau,
R. W.; Pease, R. S. Gibble, R. E.; Perez, E.; Darbeau, N. R. J . Org.
Chem., submitted.
(2) (a) White, E. H.; De Pinto, J . T.; Polito, A. J .; Bauer, I.; Roswell,
D. F. J . Am. Chem. Soc. 1988, 110, 3708. (b) White, E. H.; Field, K.
W.; Hendrickson, W. H.; Dzadzic, P.; Roswell, D. F.; Paik, S.; Mullen,
P. W. J . Am. Chem. Soc. 1992, 114, 8023. (c) White, E. H.; McGirk, R.
H.; Aufdermarsh, C. A.; Tiwari, H. P.; Todd, M. J . J . Am. Chem. Soc.
1973, 95(5), 8107. (d) Darbeau, R. W.; White, E. H. J . Org. Chem. 2000,
65, 1121. (e) Kirmse, W. J .; Moench, D. Chem. Ber. 1991, 124(1), 237.
The high reactivity of these carbocations has been
demonstrated by the interception of methylene chloride
by the deaminatively generated 1-norbornyl cation with
the eventual formation of 1-chloronorbornane as the
major product.2c Additionally, the observations of signifi-
cantly lower kT/kB values (∼ 2.5 vs ∼ 6) and significantly
higher yields of the meta isomer (∼18% vs ∼6%) from
nitrosoamide thermolyses in equimolar benzene-toluene1a
10.1021/jo001458m CCC: $20.00 © 2001 American Chemical Society
Published on Web 03/28/2001

2682 J . Org. Chem., Vol. 66, No. 8, 2001
Darbeau et al.
(vs those data from standard Friedel-Crafts approaches)
are further evidence of the enhanced reactivity of these
cations.1a
Resu lts a n d Discu ssion
Dea m in a tion of N-Ben zyl-N-n itr osop iva la m id e in
Nu cleop h ile-Sa tu r a ted Meth a n ol. N-Benzyl-N-nitro-
sopivalamide (NBNNP, 1) was decomposed (eq 1) in
methanol saturated with the sodium salts of acetate and
azide ions at 25 and 70 °C. To the detection limit of the
The deaminative approach (especially via N-nitroso-
amides) to carbenium ions is an excellent alternative to
the standard carbocationation methods such as solvolyses
and the acid-catalyzed Friedel-Crafts approaches be-
cause of the following variables: high reactivity of the
electrophile, homogeneity of the reaction mixtures,
straightforward chemistry, and excellent product bal-
ance.1b,2b,d
When the cation, e.g., benzyl cation, is stable and
devoid of removable â-hydrogens, SN1-type reaction with
available nucleophiles, e.g., the counterion and the
solvent, is its only fate. Cation-solvent reactions must
occur before diffusion of the nitrogenous molecule from
between the ions in the solvent cage as the latter process
leads to internal collapse of the ion pair to give the
corresponding ester. Thus, the N2 molecule is implicated
in this context in the partitioning of the carbocation
between its nascent counterion and the medium; the
longer the N2 molecule remains in the pocket between
the ions (prior to diffusion) the more opportunity is
afforded the carbenium ion to interact with the medium.
The yield of SDPs is increased by conditions that decrease
the diffusion rate of the nitrogenous molecule.1b,3a In-
creased solvent yields are also achieved through use of
large mole fractions of highly nucleophilic solvents,1b-d,f,2a,3a
and employment of poorly nucleophilic counterions.1b-d,2a,3a
NMR spectrometer, no benzyl acetate or benzyl azide was
observed (from nucleophilic attack at the benzyl carbon)
and neither were the methylphenyl acetates or azido-
toluenes (from nucleophilic attack on the aromatic
nucleus).5 When the decompositions were performed in
the presence of NaCN, although no tolunitriles were
detected, roughly 0.05% of benzyl cyanide was observed
(δ 3.69) at 70 °C (none observed at 25 °C).
In this study, we attempted to answer the following
questions regarding the SN1 reactions of deaminatively
generated cations:
That benzyl nitrile is detectable only at elevated
temperatures indicates a weak temperature dependence
of the product distribution ostensibly related to the
higher [CN-] at 70 °C.4d It should be borne in mind,
though, that N2 diffusion from between the ions is
accelerated at elevated temperatures resulting in less
opportunity for cation-medium interaction.1b Thus, the
rise in temperature leads to two diametrically opposed
effects on cation-cyanide interaction. An inescapable
observation, however, is that even at the elevated tem-
peratures the yields of cation-solute products are ex-
ceedingly low.5
Evidently, even in the presence of powerful nucleo-
philes such as azide (vide infra) and cyanide, no signifi-
cant substitution of the deaminatively generated benzyl
cation is observed. It is interesting that the presence of
highly reactive cations and aggressive nucleophiles in the
same reaction do not lead to appreciable yields of product
from their interactions. There would appear to be two
interpretations: (1) the benzyl cation is not the true
active intermediate in these reactions or (2) factors
conspire to prevent meaningful interactions between the
cation and the added anions.
Given the considerable body of work1,2 (vide supra) in
support of carbocation intermediacy in nitrosoamide
thermolyses, option 1 would appear to be unlikely.
J encks6a has argued that intermediates exist if and only
if at least two significant barriers for collapse of the
putative intermediate complex exist. In the case of the
deaminatively generated benzyl cation, the reverse step,
rediazoniation of the cation with molecular nitrogen, does
not occur at ambient pressures.6b indicating a significant
1. To what extent are nonsolvent nucleophiles external
to the cage capable of competitive interception of deami-
natively generated carbocations?
2. Does the solvent affect the ability of dissolved
nucleophiles to intercept the cation?
3. How do reaction temperature and the nucleophilicity
of the attacking species affect cation interception?
The source of the benzyl cation was deaminative since
(1) the purpose of the study was to determine the
vulnerability of deaminatively generated cations to attack
by nucleophilic solute and (2) because cations generated
in this manner probably represent the most reactive ones
accessible in solution-phase chemistry under ambient
conditions.1,2 In the present study, some deaminations
were performed in methanol (n ) 0.0),4a whereas others
were done solvent free (vide infra). The added nucleo-
philes were acetate (n ) 4.3),4a azide (n ) 5.8),4a and
cyanide (n ) 6.7).4a
(3) (a) The yield of SDP also rises with decreasing temperature and
to a lesser extent as the nitrogenous molecule is changed from N2 to
N2O.1b (b) Although benzyl cyanide could conceivably arise from
cyanide attack on NBNNP, the detection of products from nuclear
attack by cyanide would require the formation of the discrete benzyl
cation. Since the yields of benzyl cyanide and the tolunitriles appear
to be of the same order of magnitude, and tolunitrile formation is
impossible from NBNNP, it would appear reasonable that most, if not
all, benzyl cyanide observed from these runs must have arisen from
cyanide-cation interaction. Cyanide attack on the benzyl diazonium
ion is also ruled out based on stereochemical and steric studies.2b
(4) (a) Swain, C. G.; Scott, C. B. J . Am. Chem. Soc. 1953, 75, 141.
(b) The reduction of ion nucleophilicity by solvation may not be true
for the azide ion, since this ion is known4c to react at the diffusion
limit with short-lived cations even in strongly solvating solvents. (c)
Ritchie, C. D. Acc. Chem. Res. 1972, 5, 348. (d) In contrast, however,
Henstock found the solubility of NaCN in MeOH to be depressed at
the boiling point of the solvent (Henstock, H. J . Chem. Soc. 1934, 1340).
(5) Darbeau, R. W. Manuscript in preparation.

Deaminatively Generated Benzyl Cations
J . Org. Chem., Vol. 66, No. 8, 2001 2683
barrier for this reverse step. The physical presence of the
nitrogen molecule between the gegenions introduces a
significant steric barrier to ion-pair collapse1,2 as il-
lustrated by the observations that solvent-derived prod-
uct can predominate over ester.1c,2a,c The conclusion then
is that benzyl cations can be reasonably assumed to be
present along the reaction pathway as the active inter-
mediates. To the extent that this is true then it would
seem that under the present reaction conditions the
benzyl cation is unable to interact to any significant
extent with the added nucleophilic anions.
Ion s. NBNNP was separately saturated with sodium
acetate and sodium azide and was allowed to decompose
with stirring at 25 and 80 °C.7b After the reaction was
complete, the filtered solution was analyzed by 1H NMR.
No interception of the benzyl cation by these salts
occurred as evidenced by the absence of any benzyl
signals (apart from that due to benzyl pivalate at δ 5.28)
in the region δ 3.5-6.08 or any acetyl signals (for benzyl
acetate) at δ ∼2.4.8
Solven t-F r ee Dea m in a tion of N-Ben zyl-N-n itr oso-
p iva la m id e in th e P r esen ce of Cya n id e. NBNNP was
saturated with NaCN at 25 and 80 °C, and the suspen-
sion was vigorously stirred for 6 h and for 30 min,
respectively. At the end of the reaction, the oil was
Two plausible scenarios for the poor showing of the
anionic nucleophiles are as follows. (1) The concentrations
of the nucleophiles were too low for the ions to enter into
1
filtered and the filtrate was analyzed by H NMR. From
detectable bimolecular reaction with the benzyl cation
the runs at 25 °C, benzyl pivalate (δ 5.2,8 1.27; 99.9%),
benzyl cyanide (δ 3.69;8 trace), o-tolunitrile (δ 2.39,8
trace), and p-tolunitrile (δ 2.35,8 trace) were observed.
From the 80 °C runs the approximate product distribu-
tion was benzyl pivalate (99.7%), benzyl cyanide (∼0.2%),
o-tolunitrile (∼0.07%), and p-tolunitrile (∼0.03%). The
presence of the nitriles (and the absence of m-tolunitrile)
was confirmed by GC (vide infra).
([CN-]70°C ) 2.2 M; [-OAc]70°C ) 2.8 M; [N3
]
) 1.1
-
70°C
M). This interpretation would be consistent with the high
reactivity and short lifetimes2b (vide infra) of deamina-
tively generated carbocations, which are likely to react
in an essentially statistical manner with available nu-
cleophiles. (2) The effective nucleophilicity of the ions is
reduced in the presence of the hydrogen-bonding MeOH4b,c
to the point where their ability to significantly interact
with the benzyl cation is severely compromised.
There are several notable features of these results: (1)
solvated cyanide ions are capable of intercepting the
benzyl cations; (2) such interception occurs principally
at the benzylic position but minor attack occurs at both
the ortho and para positions;3b,5 (3) despite its high
nucleophilicity, cyanide ions compete very poorly against
the much less nucleophilic pivalate ions for the deami-
natively generated benzyl cation; (4) minimal nuclear
attack by cyanide ions on the benzyl cation occurs5 even
when these ions are generated in the absence of solvent;
(5) the extent of nuclear attack rises with increasing
temperature.5
Nucleophilic interception of the deaminatively gener-
ated benzyl cation at the aromatic nucleus is apparently
a novel observation. Indeed, interception of the aromatic
ring of benzyl and related cations by nucleophiles of any
type appears to be a rare event.5
The poor showing of cyanide is discussed in the
following sections in terms of cage and lifetime effects in
the NESIP (vide infra). The dependence upon tempera-
ture is probably related to the enhanced solubility of
sodium cyanide at elevated temperatures and to the
availability of increased energy under these conditions
for the more difficult reactions on the aromatic nucleus
(vide supra).
Lim ita tion s on th e “F r eed om ” of Dea m in a tively
Gen er a ted Ben zyl Ca tion s: Ca ge Effects in Dea m i-
n a tion . The solvent-free decomposition of N-nitroso-
amides in the presence of nucleophilic salts was inves-
tigated with two goals in mind: (1) investigation of a
novel synthetic approach to nucleophile-substituted tolu-
enes and to synthesis of those toluenes that may be
difficult to prepare via conventional aromatic substitu-
tions and (2) for studying physical organic aspects of
reactions of benzyl cations with nucleophiles with mini-
Deaminatively generated benzyl cations have been
shown to be effective at capturing even weakly nucleo-
philic solvents (e.g., benzene),1a,b and hence, an investiga-
tion of the optimal conditions for their “interceptibility”
by solute anions would require use of solvents that are
inert to the nitrosoamide, cation, and the solute but that
are sufficiently polar to dissolve appreciable amounts of
salts.
Solvents such as SO2 or the sulfonyl halides, etc., are
poor candidates because these media are inherently
acidic and would catalyze the decomposition of the
nitrosoamides, introducing an unwanted element in the
reactions. Additionally, the sulfonyl halides and other
traditional solvents such as methylene chloride and
chloroform, etc., are also good enough electrophiles to
react with cyanide and azide, etc. The question of
solubility was also an issue. There is also sufficient
precedent1d,f for the probability of the interception of the
benzyl cation by the sulfonyl species.
Fortunately, because of the facility of dediazoniation,
the deaminative approach to carbenium ions does not
require solvents for ionization to occur,1,2 and because the
N-nitrosoamides are modestly polar liquids, they are
capable of dissolving appreciable amounts of ionic com-
pounds. In any event, the presence of diluents compro-
mises the efficacy of the carbenium ion in trapping active
nucleophilic solvent.1c Ostensibly then, the best op-
portunity to engineer the interception of external anionic
nucleophiles would appear to be in the absence of
solvents.7a Thus, the decompositions of N-benzyl-N-
nitrosoamides in the presence of added nucleophiles in
the absence of solvents was investigated.7a
Solven t-F r ee Dea m in a tion of N-Ben zyl-N-n itr oso-
p iva la m id e in th e P r esen ce of Aceta te a n d Azid e
(7) (a) There are several precedents for solvent-free organic chem-
istry (e.g., that employing microwave technology, or Friedel-Crafts
reactions of liquid aromatics) when traditional solvents are either
unnecessary or unwanted. Another example is the reaction of halide
ions in molten tetrapentylammonium salts (Gordon, M.; Varughese,
P. Chem. Commun. 1971, 1160). (b) A previous study2d has shown the
mechanism of deamination to be the same at 25 and 80 °C.
(8) Pouchert, C. J . The Aldrich Library of NMR Spectra; The Aldrich
Chemical Co.: Milwaukee, WI, 1998; Vol. I.
(6) (a) J encks, W. P. Acc. Chem. Res. 1980, 13(6), 161. (b) Szele, I.;
Zollinger, H. Top. Curr. Chem. 1983, 112, 1. (c) Finneman, J . I.;
Fishbein, J . C. J . Am. Chem. Soc. 1995, 117, 4228. (d) Richard, J . P.
Tetrahedron 1995, 51, 1535. (e) McClelland, R. A.; Chan, C.; Cozens,
F.; Modro, A.; Steenken, S. Angew. Chem. Int. Ed. Engl. 1991, 30, 1337.
(f) Pezacki, J . P.; Shukla, D.; Lusztyk, J .; Warkentin, J . J . Am. Chem.
Soc. 1999, 121, 6589. (g) Amyes, T. L.; Richard, J . P. J . Am. Chem.
Soc. 1990, 112, 9507.

2684 J . Org. Chem., Vol. 66, No. 8, 2001
Darbeau et al.
solute).6f By contrast, the benzyl cation would be expected
to show less selectivity between nucleophiles resulting
in appreciable amounts of varied nucleophile-derived
benzyl species when an assortment of nucleophiles
compete for the benzyl cation.6e
Ta ble 1. Solu bility (M) of Selected Sod iu m Sa lts in
Meth a n ol a t 25 a n d 70 °C
solubility
salt
25 °C
70 °C
NaCN
NaN3
NaOAc
0.7
0.5
2.0
2.2
1.1
2.6
That solute-derived products are virtually absent from
the reaction in the present case is instructive and
consistent with the cations’ fleeting existence. Lifetimes
of 10-12-10-14 s (i.e., around the time for a bond vibra-
tion) have been determined for substituted benzyl cations
from solvolyses of substituted benzyl chlorides in water-
trifluoroethanol mixtures;6g lifetimes of deaminatively
generated 2-propyl, cyclobutonium, and 2-adamantyl and
other (nonresonance stabilized) ions have been clocked
at ∼10-10-10-9 s in similar solvents.6f It is apparent,
therefore, that the lifetime of the deaminatively gener-
ated carbocation is so short that it reacts with the more
abundant and favorably positioned solvent molecules and
nascent counterion before it can diffuse to encounter
added solutes. Thus, the virtual specificity of the deami-
natively generated benzyl cation for its nascent counter-
ion and the solvent is a boundary condition imposed upon
it by its short lifetime and ostensibly heightened by both
the cage in which the cation is generated and the
statistical limitation on the availability of solute ions.
Apart from the unlikely diffusion of the short-lived
carbocation (or any intermediate preceding it) into the
medium and chance collision with added solute, there
remains one other option for cation-solute interaction.
Ostensibly, ion-dipole interactions between the solute
and the polar nitrosoamide may exist and persist during
the nitrosoamide’s decomposition. Hence, at the moment
of formation of the cation, the solute ions are present in
the solvent cage (originally housing the nitrosoamide and
the adventitious solute) and so positioned for interaction
with the cation. To the extent that these preassociation
complexes are responsible for solute-derived products
then low product yields would be obtained because the
association constants for the formation of substrate-solute
complex are likely to be small and the benzyl cation is
not a “free” intermediate to the extent that its lifetime
is too short to allow for diffusion through the medium.6a,c,g
A F in a l Note on Azid e-Ca r boca tion Non in ter -
a ction . A significant body of information on the nature
and properties of carbocations in intimate- and solvent-
separated ion pairs has been obtained from studies in
which carbocationation has occurred in the presence of
dissolved azide ions.9 These data were gleaned from
studies determining the yield and stereochemistry of
alkyl azides derived from azide-carbocation reaction. It
is worth mentioning here that the use of azide or other
nucleophilic salts in deaminative systems would evi-
dently yield limited information because of the inability
of the deaminative carbocations to interact with these
ions in a meaningful fashion.
mal solvent perturbations. The observations, however,
show that the deaminative approach is also unsuitable
for generating appreciable yields of products from the
interaction of benzyl cations and external, nonsolvent
nucleophiles.
It would appear that the carbenium ions generated by
deamination, though essentially “uncomplexed” on for-
mation,1,2 are not “free” to pursue the theoretical range
of reactions with all available nucleophiles. There appear
to be only two species with significant opportunity to
intercept these carbenium ions: (1) the counterion, which
is trapped with it in the cage and which from a previous
study1c leads to a minimum of 20% product, and (2) the
solvent molecules that form the cage.1,2 Even in solutions
saturated with them, highly nucleophilic ions do not
compete well against either of these favored nucleophiles.
This behavior ostensibly arises due to the statistical
limitations imposed on nonsolvent, external nucleophiles
() solutes) because of their finite solubilities in the
solvent of choice (traditional or nitrosoamide). Thus, the
highly reactive deaminatively generated carbocation es-
sentially reacts in a statistical manner with the nucleo-
philes to which it is exposed during its brief insulation
from its counterion. The preponderance of solvent mol-
ecules over other nucleophiles in the medium places a
cap on the ability of the latter (no matter how high its
nucleophilicity) to intercept the cation. There also ap-
pears to be threshold nucleophilicity/concentration below
which solute-derived products are not observed. Thus, in
the competition vs solvent (when present) and counterion,
cyanide is probably favored over azide because of the
larger concentration of its sodium salt at saturation and
its inherently greater nucleophilicity. Interestingly, al-
though [AcO-] is larger than [CN-] at saturation (Table
1), its lower nucleophilicity, which is likely even further
compromised by extensive hydrogen bonding to MeOH
or by strong ion-dipole bonds to the N-nitroso moiety
(when NBNNP is solvent), must be responsible for its
nonappearance in the benzyl products.
It should be pointed out, however, that if the added
nucleophile is miscible with the solvent (e.g., D2O with
CD3CN1d) the yield of products due to this nucleophile is
proportional to its mole fraction,1d indicative of the
statistical nature of the reactions of these carbenium ions.
A Note on Lifetim e-Im p osed Lim ita tion s on Ca r -
boca tion Selectivity. An impressive body of work exists
on the lifetimes of deaminatively and solvolytically
generated carbocations.6d-g These studies indicate that,
whereas the methyl cation may react with nucleophiles
from within an encounter complex without traversing an
energy barrier, as substitution of the methyl cation by
electron releasing groups occurs, true barriers arise.6d
These barriers for carbocation-nucleophile combination
reactions are much smaller for inductively stabilized ions
than for resonance stabilized ones such as the benzyl
cation.6e Further, a similarity in the reaction barriers for
inductively stabilized cations results in a lack of selectiv-
ity of such cations to a variety of nucleophiles (solvent +
Su m m a r y
The present, novel observation of low yields of solute-
derived product from deaminatively generated carbo-
cations and the previously observed statistical intercep-
tion of D2O are consistent with the short lifetimes of
(9) (a) Richard, J . P.; Rothenberg, M. E.; J encks, W. P. J . Am. Chem.
Soc. 1984, 106, 1361. (b) Richard, J . P.; J encks, W. P. J . Am. Chem.
Soc. 1984, 106, 1373. (c) Richard, J . P.; J encks, W. P. J . Am. Chem.
Soc. 1984, 106, 1383.

Deaminatively Generated Benzyl Cations
J . Org. Chem., Vol. 66, No. 8, 2001 2685
NESIPs,1,2 and the superior positioning of the native
gegenion with respect to the deaminatively generated
carbenium ion.1,2 It would appear that the yields of
products from these reactions are controlled by the short
lifetime of these carbocations as well as cage effects in
deamination. They do, however, force the reevaluation
of the term “essentially free carbenium ions” that has
been used1,2 in connection with these ions since it is clear
from the present study that the ions though relatively
uncomplexed initially, are “fated” to appreciable reaction
only with counterion and nucleophilic solvent and are
thus not “free” in this regard.
Ostensibly, the carbocation enters into bimolecular
reaction with solvent at a rate that is significantly larger
than diffusional encounter with nucleophilic solute. Thus
solute-cation adducts can only form by virtue of solutes
present in a preassociation complex between the added
solute and the starting nitrosoamide. These complexes
are likely to be transient and present in small amounts
and hence very small yields of solute-derived products
are observed.
yellow oil. The synthesis and isolation were performed in the
1
dark: IR (neat) 1720, 1605, 1502, 1390, 1375 cm-1; H NMR
(CD3CN) δ 1.45 (s, 9H), 4.97 (s, 2H), 7.05-7.40 (m, 5H); UV
(CH2Cl2) λmax 275 nm (ꢀ ) 500), 400 nm (ꢀ ) 63) 394 nm (sh),
422 nm (ꢀ ) 66).
Decom p osition of N-Ben zyl-N-n itr osop iva la m id e (1):
(a ) In Sa tu r a ted Meth a n olic Solu tion . In an Opticlear vial
equipped with a stir bar was dissolved a calculated excess of
the appropriate sodium salt (acetate, azide, or cyanide) in 500
µL of methanol until solid remained. Approximately 50 mg of
1 was then added with stirring, and the vial was closed and
incubated in the dark at the desired temperature until
decomposition was complete (overnight at 25 °C or 4 h at 70
°C). The solution was then filtered and treated as described
above. For the acetate and azide salts at both temperatures
and for NaCN at 25 °C, the 1H NMR spectra showed only
benzyl pivalate and benzyl methyl ether in 1:3 ratio as above.
However, for saturated methanolic NaCN, a new signal at δ
3.69 was observed. The sample was spiked with commercial
benzyl cyanide when selective growth of the δ 3.69 signal was
observed. (b) In th e Absen ce of Solven t. These runs were
set up in the same fashion as just described except that ∼250
mg of 1 was used as solvent and reagent. At the end of the
runs, the samples were filtered, an aliquot was dissolved in
1
CDCl3, and H NMR spectra were taken. For the acetate and
1
azide salts at both temperatures, the H NMR spectra showed
Exp er im en ta l Section
only benzyl pivalate. However, for saturated NaCN, benzyl
pivalate (δ 5.2,8 1.27; 99.9%), benzyl cyanide (δ 3.69,8 trace),
o-tolunitrile (δ 2.39,8 trace), and p-tolunitrile (δ 2.35,8 trace)
were observed from the runs at 25 °C. From the 80 °C runs
the product distribution was benzyl pivalate (99.7%), benzyl
cyanide (∼0.2%), o-tolunitrile (∼0.07%), and p-tolunitrile
(∼0.03%).
Ma t er ia ls a n d Met h od s. All commercial reagents were
reagent grade and were used without further purification.
Spectra were recorded on 300 MHz FT-NMR, FT-IR, and UV-
vis spectrometers. GC data were obtained using an HP-5 cross-
linked 5% Ph Me capillary column (30 m × 0.25 mm × 0.25
µm) at 150 °C, 8.1 psi (carrier gas ) He; FID detector).
Sta bility of th e P r ecu r sor s: Ha n d lin g a n d Stor a ge.
N-Benzyl-N-nitrosopivalamide in addition to being photolabile
and thermolabile is unstable in the presence of acids, bases,
and moisture. As a result, the dry, neutral oil was stored in a
desiccator in liquid N2. The compounds used in this study could
be stored in this fashion indefinitely. The half-life of NBNNP
is ∼15 min at 25 °C in MeOH. All compounds were handled
in the dark. Ca u tion ! N-Nitrosoamides should be handled
with extreme care because of their possible mutagenicity10a
and carcinogenicity (local and systemic).10b Efficient fume
hoods and appropriate personal protection (chemical-resistant
gloves, safety glasses, lab coat, etc.) are recommended when
handling these compounds.
Deter m in a tion of Solu bility of Sod iu m Sa lts in Meth a -
n ol. Aliquots of a known excess mass of the appropriate
anhydrous salt were added in portions with stirring to
methanol (10 mL) in an Erlenmeyer flask in water baths at
25 and 70 °C. After saturation (as evidenced by the appearance
of undissolved solid), a heated pasteur pipet with a cotton wool
plug at the end was used to remove the methanolic solution,
and the solvent volume was measured. The mass of the added
salt was determined by the difference between that of the
original excess and that of the final (“unadded”) mass. The
undissolved residue in the flask was weighed and its mass was
used to correct that required for saturation. From the volume
of methanolic solution and the corrected mass of added solute,
the solubilities of the salts in g mL-1 were determined
(Table 1).
N-Ben zylp iva la m id e was prepared from the method of
Heyns and von Bebenburg:11a mp 81-82 °C (lit.11b mp 81-82
°C); IR (KBr) 3309, 1689, 1510, 1390, 1375 cm-1 1H NMR
;
Ga s Ch r om a togr a p h ic Detection of th e Isom er ic Tolu -
n itr iles a n d Ben zyl Cya n id e. Aliquots of product mixtures
from the reactions in saturated NaCN were diluted in ether
and determined by gas chromatography. Signals at tR ) 5.0,
5.3, and 5.8 min corresponding to the retention times of the
commercial o/ p-tolunitriles and benzyl cyanide, respectively,
were observed. No signal at tR ) 5.2, for m-tolunitrile, was
detected. The yields of the nitriles were very low, and no
reproducible ratios could be determined (although the com-
pounds were always detected above the baseline noise).
(CDCl3) δ1.27 (s, 9H), 4.44 (d, 2H, J ) 7 Hz), 5.90 (bs, 1H),
7.26-7.32 (m, 5H); UV (ET2O) λmax ) 284 nm (ꢀ ) 209).
N-Ben zyl-N-n itr osop iva la m id e (1). A mixture of N-
benzylpivalamide (95.5 mg, 0.5 mmol), NaOAc (250 mg, 3.0
mmol), and Na2SO4 (0.5 g) was dried at oil pump vacuum.
Methylene chloride (3.0 mL), freshly distilled from P2O5, was
added to the solid material (under N2), and the suspension
was cooled to -78 °C. A solution of N2O4(l) (0.2 mL, 3.1 mmol)
in CH2Cl2 (1.0 mL) at -78 °C was then added to the stirred
suspension at -78 °C, which was then allowed to warm to -25
°C over 10 min. After a further 15 min at -25 °C, the
suspension was evaporated in vacuo for ∼15 min until a lemon
yellow color was observed. Ether at -20 °C was then added,
and the suspension was washed in turn with saturated
solutions of NaCl, Na2CO3, and NaCl at -5 °C. The organic
phase was dried over Na2SO4 at -30 °C and then evaporated
in vacuo (at -30 °C) to yield 0.11 g (0.5 mmol, 100%) of a lemon
Ack n ow led gm en t is made to the Calcasieu Parish
Industrial and Development Board Endowed Professor-
ship, administered by the McNeese State University
Foundation, and the Shearman Research Initiative,
MSU and to the Department of Chemistry, MSU, for
partial support of this research. We also acknowledge
the American Chemical Society, in general, and the
SWLA ACS chapter, in particular, for support of the
Project SEED students. We also extend our gratitude
to Ms. J oan E. Vallee and Dr. Mark S. Delaney.
(10) (a) Lee, K.; Gold, B.; Mirvish, S. Mutat. Res. 1977, 48, 131. (b)
Preussman, R.; Stewart, B. W. Chemical Carcinogenesis; Searle, C.,
Ed.; ACS Monograph No. 182; American Chemical Society: Washing-
ton, DC, 1984; pp 643-828.
(11) (a) Heyns, K.; V. Bebenburg, W. Chem. Ber. 1953, 86, 278. (b)
Beilstein Vol. 12, 3rd Suppl., p 2346.
J O001458M