Acidity of Di- and triprotected hydrazine derivatives in dimethyl sulfoxide and aspects of their alkylation

Article abstract of DOI:10.1021/jo050680u

The pK_a values in DMSO for 22 di- and triprotected hydrazine NH acids and two monosubstituted hydrazines have been determined using potentiometric titration. The results of density functional theory calculations at the B3LYP/6-311+G** level of gas-phase acidities of a representative selection of mono-, di-, and trisubstituted hydrazines are compared with both the relevant published and novel experimental titration data. In the course of this work, a rough estimation of the pK_a value of hydrazine in DMSO (ca. 38.0) has been deduced. For typical triprotected compounds of this kind containing moderately electron-withdrawing carbamate and imidodicarbonate or arenesulfonyl-carbamate functions the pK_a values fall in the range 15.1-17.3, whereas for N,N′-diprotected hydrazines with a carbamate and an aromatic sulfonyl group the corresponding values are 12.7-14.5. Several of these triprotected derivatives have recently been applied preparatively in stepwise synthesis of substituted hydrazines using alkyl halides as electrophiles in the presence of a phase transfer catalyst, and a few of them, with varying success, have been examined in model experiments with benzyl alcohol, triphenylphosphine, and diethyl azodicarboxylate in the Mitsunobu reaction. The dependence of the reactivity on the intrinsic acidity of the hydrazines in this reaction is highlighted. Furthermore, the regioselective alkylation of an N,N′-diprotected hydrazine can be rationalized on the basis of the presented data.

Full text of DOI:10.1021/jo050680u

Acidity of Di- and Triprotected Hydrazine Derivatives in Dimethyl
Sulfoxide and Aspects of Their Alkylation
Ulf Ragnarsson,*,† Leif Grehn,† Juta Koppel,‡ Olavi Loog,§ Olga Tsˇubrik,§ Aleksei Bredikhin,§
Uno Ma¨eorg,§ and Ilmar Koppel*,‡
Department of Biochemistry, University of Uppsala, Biomedical Center, P.O. Box 576, SE-751 23
Uppsala, Sweden, Institute of Chemical Physics, Department of Chemistry, and Institute of Organic and
Bioorganic Chemistry, University of Tartu, Jakobi 2, Tartu 51014, Estonia
urbki@bmc.uu.se; ilmar@chem.ut.ee
Received April 6, 2005
The pKa values in DMSO for 22 di- and triprotected hydrazine NH acids and two monosubstituted
hydrazines have been determined using potentiometric titration. The results of density functional
theory calculations at the B3LYP/6-311+G** level of gas-phase acidities of a representative selection
of mono-, di-, and trisubstituted hydrazines are compared with both the relevant published and
novel experimental titration data. In the course of this work, a rough estimation of the pKa value
of hydrazine in DMSO (ca. 38.0) has been deduced. For typical triprotected compounds of this kind
containing moderately electron-withdrawing carbamate and imidodicarbonate or arenesulfonyl-
carbamate functions the pKa values fall in the range 15.1-17.3, whereas for N,N′-diprotected
hydrazines with a carbamate and an aromatic sulfonyl group the corresponding values are 12.7-
14.5. Several of these triprotected derivatives have recently been applied preparatively in stepwise
synthesis of substituted hydrazines using alkyl halides as electrophiles in the presence of a phase
transfer catalyst, and a few of them, with varying success, have been examined in model experiments
with benzyl alcohol, triphenylphosphine, and diethyl azodicarboxylate in the Mitsunobu reaction.
The dependence of the reactivity on the intrinsic acidity of the hydrazines in this reaction is
highlighted. Furthermore, the regioselective alkylation of an N,N′-diprotected hydrazine can be
rationalized on the basis of the presented data.
Introduction
presents considerable experimental difficulties, which are
particularly pronounced in the preparation of alkylated
derivatives on a laboratory scale, where, as a rule,
mixtures of starting material, the desired product, and
overalkylated hydrazines are obtained. These difficulties
could, in principle, be reduced by application of suitable
amino-protecting groups.3 Some time ago it occurred to
us that triprotected hydrazine derivatives4 would allow
substitutions to be driven to completion without forma-
tion of multisubstituted products, and a prototypic re-
Substituted hydrazines are nowadays of great technical
and commercial importance,1 and a large number of
methods for their synthesis have been developed.2 Direct
substitution of one or more hydrazine hydrogens often
* To whom correspondence should be addressed. Phone: +46 18 471
45 55. Fax: +46 18 55 21 39 (U.R.). Fax: +372 7 375 264 (I.K.).
† University of Uppsala.
‡ Institute of Chemical Physics, University of Tartu.
§ Institute of Organic and Bioorganic Chemistry, University of
Tartu.
(1) Hydrazine and Its Derivatives. In Kirk-Othmer Encyclopedia of
Chemical Technology, 4th ed.; Wiley: New York, 1995; Vol. 13.
(2) (a) Comprehensive review: Jensen-Korte, U.; Mueller, N. In
Methoden der Organischen Chemie (Houben-Weyl), 4th ed.; Klamann,
D., Ed.; Thieme: Stuttgart, 1990; Vol. 16a, pp 425-648. (b) Recent
review: Huddleston, P. R.; Coutts, I. G. G. In Comprehensive Organic
Functional Group Transformations; Katritzky, A. R., Meth-Cohn, O.,
Rees, C. W., Eds.; Pergamon/Elsevier: Oxford, 1995; Vol. 2, pp 371-
383. (c) Selective review: Ragnarsson, U. Chem. Soc. Rev. 2001, 30,
205.
(3) (a) Greene, T. W.; Wuts, P. G. M. Protective Groups in Organic
Synthesis, 3rd ed.; Wiley-Interscience: New York, 1999; Chapter 7.
(b) Kocien´ski, P. J. Protecting Groups; Thieme: Stuttgart, 1994;
Chapter 6.
(4) (a) Ma¨eorg, U.; Grehn, L.; Ragnarsson, U. Angew. Chem. 1996,
108, 2802; Angew. Chem., Int. Ed. Engl. 1996, 35, 2626. (b) Grehn, L.;
Lo¨nn, H.; Ragnarsson, U. Chem. Commun. 1997, 1381. (c) Ragnarsson,
U.; Grehn, L.; Maia, H. L. S.; Monteiro, L. S. J. Chem. Res., Synop.
2000, 6; J. Chem. Res., Miniprint 2000, 128.
10.1021/jo050680u CCC: $30.25 © 2005 American Chemical Society
Published on Web 06/28/2005
5916
J. Org. Chem. 2005, 70, 5916-5921

Acidity of Di- and Triprotected Hydrazines
TABLE 1. Acidity of Hydrazine Derivativesa in DMSO (pKa) and in the Gas Phase (GA)
Part A. Based on Published pKa(DMSO) Data
Part B. Di- and Triprotected Derivatives
GAb
this work
GAb/
g
entry
P2
P3
P4
ref
pKa
entry
P2
P3
P4
source
pKa
categoryh
H
H
H
H
H
H
H
H
H
H
H
Ph
Ph
Ph
Me
Me
H
-
-
391.7
1
Ts
Ts
1-Ns
2-Ns
Cbs
Z
Boc
Z
Z
Z
Boc
TFA
Boc
CH3CO
Boc
Ts
Boc
Z
9a
15.5
16.0
15.1
15.4
12.6
14.0
14.5
13.4
13.6
12.7
A
1
H
Ph
8c
28.8c 354.9d/370.2e
2
Boc Boc
4c
A
2
H
CH3CO 8b 21.8c 350.7
3
Boc
Boc
Boc
H
Z
4c
4c
A
3
H
PhCO
CO2Et
PhSO2
Ph
CH3CO 8c
CH3CO 8c
8b 18.9c 341.0
4
Z
A
4
H
8c
22.2c 354.4
5
Z
4b,c
9a
4c
9b
9b
4b,c
A
5
H
8b 17.1c 339.1
8b 26.2c 343.4
6
Ts
B
6
Ph
Ph
7
H
Ts
B
7
18.4
344.6
8
H
1-Ns
2-Ns
Cbs
TFA
TFA
Tf
B
8
CH3CO
PhCO
Ph
16.7c 332.5
9
H
B
9
PhCO
Ph
8c
8c
8c
8c
13.7
24.5
12.1
17.9
326.3
-
-
-
340.0
337.7
322.5
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
H
B
10
11
12
13
14
15
16
17
18
19
20
21
H
this work 10.0
21
12
318.2/B
Ph
DNP
PhCO
PhCO
PhSO2
H
H
7.4
8.2
304.8/B
Ph
H
315.3/B
+
Me
8d 19.7
8b 15.8
H
PPh3+ this work 11.4
B
B
B
C
C
C
C
C
C
Me
H
PPh3
H
9c
9d
9.9
maleic hydrazide
8c
8c
-
-
-
-
-
13.2
Boc
15.9
17.3
17.1
17.1
17.8
16.9
17.1
phthalic hydrazide
H
12.7c 323.9
12.0f 321.8
Boc
Z
Boc Boc
Boc Boc
4a,9e
9e,f
9e
H
H
Tf
Tf
H
Tf
4.6f
0.4f
4.3f
300.4
288.2
299.5
(CH2)2CN Boc Boc
Tf
Tf
Tf
Ph
Ph
Ph
H
Boc Boc
Boc
H
H
H
9g
TFA
TFA
TFA
H
TFA
TFA
Z
Z
TFA
PPh3
9g
17.0f 336.5
13
this work 16.6c,i 330.7
+
H
14
14.3
-
a For labeling of substituents, see Scheme 1. Abbreviations: DNP, 2,4-dinitrophenyl, Boc, tert-butoxycarbonyl; Cbs, 4-cyanobenzene-
sulfonyl; 1- and 2-Ns, 1- and 2-naphthalenesulfonyl; PPh3+, triphenylphosphonium; Tf, trifluoromethanesulfonyl; TFA, trifluoroacetyl;
Ts, tosyl; Z, benzyloxycarbonyl. b Calculated in this work at the DFT/B3LYP 6-311+G** level, in kcal/mol units. c Hydrazine used to
derive eq 3. d Deprotonation in the R-position to the phenyl ring. e Deprotonation in the â-position to the phenyl ring. f pKa values calculated
from eq 2. g Measured in this work. h Based on their pKa values the di- and triprotected derivatives were assigned to category A, B, or C
as discussed below. i The pKa of CF3CONH2 (17.2) is given in ref 8b.
SCHEME 1. Stepwise Synthesis of
Tetrasubstituted Hydrazines from a Triprotected
Reagent
In a previous article, it was demonstrated that for
imidodicarbonates and tosylcarbamates there is a clear
connection between their pKa in DMSO solution and the
yield of product with a selected alcohol in the Mitsunobu
reaction.7 Therefore, to understand and rationalize the
behavior of these protected hydrazines upon alkylation
with alcohols in this way, it would obviously be helpful
to know their pKa values in this solvent, in particular
since such values for only simple and distantly related
derivatives and not for hydrazine itself have been deter-
mined previously.8 Altogether the literature contains
more than 30 pKa values related to DMSO of aryl, acyl,
benzenesulfonyl, ethoxycarbonyl, and similar, mostly
mono- and disubstituted, predominantly acyclic hydra-
zines, hydrazides, and hydrazones in the range from 28.8
(PhNHNH2) to 12.1 (2,4-(NO2)2C6H3NHNPh2). A few
pKa’s of acylated hydrazines, including BocHNNBoc2,
were also measured in aqueous solution.5a
agent (entry 17 in Table 1, Part B) of this kind was
prepared and investigated.4a Subsequently, another re-
agent (entry 5) with three orthogonal protecting groups
was developed and used for the first synthesis of tetra-
substituted derivatives in which all substituents were
different.4b These syntheses were accomplished in a
stepwise fashion, each substitution (with RX) being
followed by a selective removal of a protecting group (P)
as shown in Scheme 1.
In this article, a number of di- and triprotected
hydrazine reagents4,9 with arenesulfonyl and alkyloxy-
carbonyl and, in three cases, triphenylphosphonium
Initially, these N-alkylations were carried out with
halides under phase transfer catalysis (PTC) conditions.4a
Subsequently, alcohols4b,c,5 have also been applied in the
first alkylation step under Mitsunobu conditions, that is,
in the presence of triphenylphosphine and an azodicar-
boxylate, which requires an acidic nitrogen component.6
(7) Koppel, I.; Koppel, J.; Degerbeck, F.; Grehn, L.; Ragnarsson, U.
J. Org. Chem. 1991, 56, 7172.
(8) (a) Bordwell, F. G.; Algrim, D. J. J. Am. Chem. Soc. 1988, 110,
2964. (b) Bordwell, F. G. Acc. Chem. Res. 1988, 21, 456. (c) Zhao, Y.;
Bordwell, F. G.; Cheng, J.-P.; Wang, D. J. Am. Chem. Soc. 1997, 119,
9125. (d) Bordwell, F. G.; Harrelson, J. A., Jr.; Lynch, T.-Y. J. Org.
Chem. 1990, 55, 3337.
(9) (a) Grehn, L.; Nyasse, B.; Ragnarsson, U. Synthesis 1997, 1429.
(b) Nyasse, B.; Grehn, L.; Maia, H. L. S.; Monteiro, L. S.; Ragnarsson,
U. J. Org. Chem. 1999, 64, 7135. (c) Tsˇubrik, O.; Ma¨eorg, U. Org. Lett.
2001, 3, 2297. (d) Grehn, L.; Ragnarsson, U. Tetrahedron 1999, 53,
4843. (e) Ma¨eorg, U.; Pehk, T.; Ragnarsson, U. Acta Chem. Scand.
1999, 53, 1127. (f) Ma¨eorg, U.; Ragnarsson, U. Tetrahedron Lett. 1998,
39, 681. (g) Loog, O.; Ma¨eorg, U.; Ragnarsson, U. Synthesis 2000, 1591.
(5) (a) Brosse, N.; Pinto, M.-F.; Jamart-Gre´goire, B. J. Org. Chem.
2000, 65, 4370. (b) Brosse, N.; Pinto, M.-F.; Bodiguel, J.; Jamart-
Gre´goire, B. J. Org. Chem. 2001, 66, 2869. (c) Pinto, M.-F.; Brosse,
N.; Jamart-Gre´goire, B. Synth. Commun. 2002, 32, 3603.
(6) (a) Mitsunobu, O. Synthesis 1981, 1. (b) Hughes, D. L. Org. React.
N.Y. 1992, 42, 335.
J. Org. Chem, Vol. 70, No. 15, 2005 5917

Ragnarsson et al.
protecting groups have been studied, and their pKa values
in DMSO (pKa(DMSO)) determined by analogy with imi-
dodicarbonates and tosylcarbamates previously.7 Also, for
the sake of comparison with these experimental values,
the gas-phase acidities (GA) of a representative selection
of 22 hydrazines with widely different structures and
pKa(DMSO) were calculated using a density functional
theory (DFT) approach at the B3LYP/6-311+G** level.10
The computations were carried out using the Gaussian
2003 series of programs. DFT calculations were per-
formed using the B3LYP hybrid functional. Full geometry
optimizations and vibrational analyses were performed
using the 6-311+G** basis set except for hydrazine, for
which also the 6-311++G** basis set with diffusion
functions on hydrogen atoms was used. This approach
has been demonstrated by some of us to describe with
reasonable accuracy both the gas-phase basicities and
acidities of a wide variety of relatively simple molecules.10
All stationary points were found to be true minima (Nimag
) 0). Unscaled B3LYP 6-311+G** frequencies were used
to calculate the GA of neutral acids AH and proton
affinities (PA(A-)) of the conjugated anionic bases (A-)
to take into account the zero-point frequencies, finite
temperature 0-298 K correction, the pressure-volume
work term, and the entropy terms as appropriate. The
terms GA and PA refer to the following equilibrium (eq
1) and are defined as follows:
and supplemented them with calculated GA data. In Part
B, the results of our pKa determinations dealing with a
number of recently prepared di- and triprotected hydra-
zine reagents (entries 1-19) and a few simple derivatives
are presented, together with some additional GA data.
It also contains such information for two monoprotected
hydrazines prepared for comparison purposes.
Inspection of Table 1 shows that, for the set of
hydrazines studied, both the calculated GA and the
experimental pKa(DMSO) vary widely, the former by ca. 104
kcal/mol corresponding to a difference in equilibrium
constants of 76 powers of ten, and the latter by more than
21 units. The data fit the linear relation (2):
GA ) a + b‚pKa(DMSO)
(2)
where a ) 287.0 ( 3.2 and b ) 2.91 ( 0.20, n ) 15, r )
0.971, and s ) 3.4
Equation 2 holds for most mono-, di-, and trisubstituted
hydrazines, including compounds with electron-acceptor
substituents, starting with 1,2-bistrifluoroacetyl hydra-
zine (pKa ) 7.4) and ending with H3CCONHNH2 and
EtO(O)CNHNH2 (pKa ) 22.2). Surprisingly, the pKa of
phenylhydrazine fits the same straight line only on
condition that the calculated GA value refers to depro-
tonation from the NH2 group (â-position to the phenyl
group) (eq 2). Relationships similar to eq 2 between the
experimental GA and pKa values in DMSO or aqueous
solution were earlier observed to hold for other classes
of NH acids (anilines, imides, amides, heterocycles, etc.).11
Equation 2 shows that upon going from the gas phase
into DMSO solution the effects of substituents on the GA
of hydrazines are rather significantly attenuated (com-
pare with ref 11). Equation 2 allows one to make rough
estimates of solution-phase acidities of some hydrazines
that are either hard to prepare (e.g., TfNHNH2) or are
predicted to be too acidic (e.g., TfNHNHTf, Tf2NNHTf,
etc.) or too weak (e.g., methylhydrazines) to be measured
directly in DMSO solution. A few such pKa(DMSO) values
are given at the end of Part A and will be discussed later
in the article.
The first five compounds in Part A illustrate the
acidifying effect of a single phenyl or acyl substituent in
hydrazine. PhSO2NHNH2 is by ca. 5 powers of 10 a
stronger acid than EtOCONHNH2, used here to estimate
the influence of a single carbamate protecting group such
as Z or Boc, for which literature data is missing. As
follows from eq 2, even much higher acidity (ca. 12) could
be expected for trifyl hydrazine. Unfortunately, all of our
attempts to synthesize that compound (compare also ref
15) have thus far been inconclusive. Upon going from
mono- to diacylated hydrazines (entries 8 and 9 in Part
A), a significant trend toward higher acidity is noticeable.
TfNHNHBoc, TFANHNHTFA, and TFANHNHBoc are
of the same type and, as can be seen, the acidifying effect
of the Tf and TFA groups is much stronger than that of
others studied. This is evident also for 1,2-bistrifyl
hydrazine, for which the calculated GA reaches that of
sulfuric acid.10b
∆Gacid, ∆Hacid
AH [GA ) ∆G
] A- + H+
(1)
PA(A-) ) ∆Hacid
acid
Most of the calculated GA values (in kcal/mol units)
alongside some representative literature pKa(DMSO) data
are given in Table 1A, whereas some directly related to
our present experimental study are found in Table 1B.
The calculated energies, enthalpies, and free energies of
the substituted hydrazines and corresponding deproto-
nated species as well as proton affinities PA(A-) are
presented in the Supporting Information.
The new solution and GA data are interpreted on a
structural basis and used to rationalize previously re-
ported work dealing with the application of such com-
pounds in the synthesis of multisubstituted hydrazines.
Alkylhydrazines and hydrazine itself are too weak NH
acids for measurement of their pKa(DMSO) due to autopro-
tolysis (pKauto ) 35 for DMSO)8b like alkylamines and
ammonia (an indirect estimate by Bordwell and Algrim8a
places the pKa(DMSO) of ammonia at 41 ( 1). Therefore,
in this article we also make an attempt to estimate the
pKa(DMSO) of hydrazine by using (1) a correlation between
the calculated gas-phase and solution (DMSO) acidities
for substituted hydrazines (eq 2 below), (2) another
correlation of pKa(DMSO) data for ammonia and hydrazine
derivatives (eq 3 below), and (3) a linear relation between
experimental gas-phase acidities and pKa(DMSO) as de-
scribed in ref 11.
Results and Discussion
In Table 1, Part A, we have collected pKa(DMSO) values
from the literature for a selection of relevant hydrazines
(10) (a) Burk, P.; Koppel, I. A.; Koppel, I.; Leito, I.; Travnikova, O.
Chem. Phys. Lett. 2000, 323, 482. (b) Koppel, I. A.; Burk, P.; Koppel,
I.; Leito, I.; Sonoda, T.; Mishima, T. J. Am. Chem. Soc. 2000, 122, 5114.
(11) Koppel, I. A.; Koppel, J.; Maria, P.-C.; Gal, J.-F.; Notario, R.;
Vlasov, V. M.; Taft, R. W. Int. J. Mass Spectrom. Ion Processes 1998,
175, 61.
(12) Hendrickson, J. B.; Sternbach, D. D. J. Org. Chem. 1975, 40,
3450.
(13) Atherton, F. R.; Lambert, R. W. Tetrahedron 1983, 39, 2599.
(14) Zimmer, H.; Singh, G. J. Org. Chem. 1963, 28, 483.
(15) Chamberlin, A. R.; Bond, F. T. J. Org. Chem. 1978, 43, 154.
5918 J. Org. Chem., Vol. 70, No. 15, 2005

Acidity of Di- and Triprotected Hydrazines
The trisubstituted derivatives (entries 10-14 in Part
A) are all too different in structure from those in Part B
to provide a meaningful comparison. Among the latter
are first some triprotected hydrazines, whose aromatic
sulfonyl residues P2 (Ts, 1-Ns, 2-Ns, Cbs; entries 1-5)
together with P3 form sulfonylcarbamates, whereas P4
are carbamate functions. These substances have pKa(DMSO)
values in the range 15.1-16.0, with the exception of the
last one. For the corresponding 1,2-diprotected sub-
stances (entries 6-10) with a hydrogen at the sulfona-
mide nitrogen, these values are 12.7-14.0, thus giving
rise to two fairly well-defined sets of hydrazine deriva-
tives, called A and B below. Although acidity in general
should be determined primarily by the substituent
directly attached to the deprotonation site, the significant
difference in this respect between EtOCONHNH2 and
category A compounds demonstrates an additional con-
siderable contribution from the electron-withdrawing
sulfonylcarbamate NP2P3 moiety. This seems to partly
offset the effect of the sulfonyl groups in type B com-
pounds, and as a result, the gap in pKa(DMSO) between
these two categories is reduced to 2-3 units. Among the
1,2-disubstituted compounds experimentally studied, two
Jamart-Gre´goire et al.5a In DMSO, literature values
range from 24.5 to 12.1 (entries10-14 in Part A), the last
value referring to a compound with the strongly electron-
withdrawing 2,4-dinitrophenyl substituent at the NH
site.
On the basis of DFT calculations and eq 2, the pKa(DMSO)
for Tf2NNHTf and (TFA)2NNHTFA can be estimated to
be around 0.4 and 4.3, respectively. The calculated GA
values of those compounds equal or exceed that of triflic
acid10 and are determined by a combination of the field
inductive effects of the substituents (Tf and NTf2 or TFA
and N(TFA)2) at the N-atoms and most strongly by
resonance stabilization of the anionic species by the
electron-accepting substituents (Tf or TFA groups) at the
ionization center. Therefore, due to the lack of such
stabilization, the acidifying effect of the NTf2 or N(TFA)2
moieties should be weaker.
The data for two of our tosyl derivatives (entries 6 and
7, Part B) should also be compared with the pKa(DMSO)
for tosyl amide, for which values of 15.74 and 16.3 have
been reported,7,17 indicating a moderately acidifying effect
of the additional substituted nitrogen. In aqueous
solution,5a the pKa values of the only three compounds
studied, two N-alkoxycarbonylaminophthalimides and
BocNHNBoc2 identical with that in entry 17, are around
11 or, more exactly, 11.3 for the last compound, that is,
ca. 6 pKa units lower than in DMSO, which is in
reasonable agreement with earlier findings for neutral
NH acids of similar structure and acid strength.7,17
Although this triprotected hydrazine reagent (entry 17)
is among the least acidic substances in our set of
compounds, it could be alkylated readily by halides under
various PTC conditions.4a,9e A related reagent (entry 18)
exhibits a similar pKa(DMSO) and could also be efficiently
alkylated in this way,9d,e although extended reaction
times should be avoided due to the increased sensitivity
to bases of its imidodicarbonate function. Since, in our
experience, sulfonylcarbamates are completely stable
toward many strong bases, procedures based on PTC
techniques were reliable for alkylation of triprotected
hydrazine reagents containing such functions. Detailed
experimental conditions have been given elsewhere.4c,9a
On the other hand, Jamart-Gre´goire et al. reported
that all attempts to alkylate BocNHNBoc2 using the
Mitsunobu protocol failed.5a Using benzyl alcohol, triph-
enylphosphine (TPP), and diethyl azodicarboxylate
(DEAD), we obtained a maximum of 7% of benzylated
product with this reagent as estimated by 1H NMR.
Similarly with the Boc2Z-reagent (entry 18), even with
considerable excess of TPP and DEAD, the yield of
1-benzyl-1,2-Boc2-2-Z-hydrazine was low (37%; not in-
cluded in the Experimental Section). These results differ
from those first obtained with a more acidic reagent4b
(entry 5) and from other ones belonging to Category A,
as shown in the Experimental Section and Supporting
Information. Simple benzyl alcohols4b,c and also 2-pro-
panol18 have been shown to react essentially quantita-
tively. We conclude that reagents of Category C, contrary
+
contain the positively charged PPh3 group (entries 14
and 15), which makes them more acidic than those with
an aromatic sulfonyl group. This indicates not only that
+
the PPh3 group has stronger electron-withdrawing
properties, but also that it can presumably better stabi-
lize the zwitterionic species formed on protolysis by
electrostatic ion-ion interaction. For the compound in
entry 15, the increase in acidity approaches five and three
pKa units as compared to those in entries 7 and 10,
respectively.16
The next compound, H2NNTsBoc (entry 16), is a rare
example of a diprotected hydrazine that has both sub-
stituents (P2 and P3) on the same nitrogen atom. Struc-
turally, it is related to the triprotected derivatives in
entries 1 and 2 with about the same acidity, which again
demonstrates the strongly acidifying effect of its sulfo-
nylcarbamate moiety. Roughly similar acidity (17.1) is
predicted via eq 2 also for (TFA)2NNH2 and Ac2NNH2 on
the basis of their calculated GA. The rather moderate
acidity of this class of compounds results from the lack
of a resonance-stabilizing electron acceptor group at the
deprotonation site.
The three following triprotected derivatives without
both sulfonyl and phenyl moieties (entries 17-19) exhibit
pKa(DMSO) values in the very narrow range 17.1-17.3
(category C). Comparing data for the two C compounds
containing only Z- and Boc-groups with those of category
A, it appears that the acidifying effect of a sulfonylcar-
bamate in relation to an imidodicarbonate moiety is less
than two pKa units. Similar low values, 16.9-17.8, are
exhibited by the three phenylhydrazines studied (entries
20-22; therefore referred to category C) with Boc-
NHNPhBoc of lowest acidity among all the compounds
in this study, spanning 8 orders of magnitude (cf. also
Part A entries 1, 7, and 12).
The only previous pKa determinations on trisubstituted
hydrazines that we know of have been performed in
DMSO by Bordwell et al.8b-d and in aqueous solution by
(17) (a) Ludwig, M.; Pytela, O.; Vecˇerˇa, M. Collect. Czech. Chem.
Commun. 1984, 49, 2593. (b) Koppel, I. A.; Koppel, J.; Leito, I.; Koppel,
I.; Mishima, M.; Yagupolskii, L. M. J. Chem. Soc., Perkin Trans. 2 2001,
229.
(18) Ragnarsson, U.; Fransson, B.; Grehn, L. J. Chem. Soc., Perkin
Trans. 1 2000, 1405.
(16) Zhang, X.-M.; Fry, A. J.; Bordwell, F. G. J. Org. Chem. 1996,
61, 4101.
J. Org. Chem, Vol. 70, No. 15, 2005 5919

Ragnarsson et al.
in both the ammonia8a,b (XNH2) and hydrazine8b,c
(XNHNH2) series the introduction of a single, polarizable,
electron-accepting nonalkyl substituent X (e.g., Ar, CH3-
CO, PhCO, COOEt, TFA, and PhSO2) has a drastic
influence on the acidity. A similar effect can be noticed
also for cyclic substituents (e.g., in phthalimide and the
respective phthalic hydrazide or symmetrically disubsti-
tuted derivatives Ph2NH and Ac2NH). The relevant
experimental data related to these two reaction series
for the six above-mentioned monosubstituted derivatives,
for disubstituted Ph- and Ac-derivatives and for phthalic
hydrazide and phthalic imide can be expressed as follows
(eq 3):
SCHEME 2. Simplified Stepwise Synthesis of
Tetrasubstituted Hydrazines from a Diprotected
Reagenta
a P3 ) Z or Boc; P4 ) ArSO2; TPP ) triphenylphosphine; DEAD
) diethyl azodicarboxylate; X ) halide.
to A reagents, are too weakly acidic to undergo smooth
Mitsunobu reaction under these conditions.
pKa(XNHNH2) )
(1.6 ( 2.6) + (0.86 ( 0.12)pKa(XNH2) (3)
Interestingly, the presented pKa data also rationalizes
the behavior of a diprotected derivative (entry 7) which
undergoes perfectly regioselective benzylation on its most
acidic nitrogen under Mitsunobu conditions (see Experi-
mental Section4c). Similar 1,2-diprotected compounds
(e.g., entries 6 and 8-10) of similar acidity might also
be attractive synthetically (Scheme 2), because, in addi-
tion to being simple to prepare, they provide a shortcut
to tri- and tetrasubstituted hydrazines by elimination of
one step, as shown in Scheme 2.
On several occasions, a third method to alkylate
protected hydrazines in addition to halides under PTC
and alcohols under Mitsunobu conditions has been at-
tempted; namely, with triflates made from alcohols. One
successful example based on BocNHNBoc2 can be found
in the Experimental Section, in which deprotonation is
accomplished by BuLi. Another simple example is given
in the Supporting Information, demonstrating use of a
triflate in the second alkylation step. Other attempts at
alkylation of hydrazines with triflates have thus far given
less satisfactory results.
Comparison of the intrinsic, gas-phase acidities of
ammonia (∆Gacid ) 396.1 kcal/mol)11 and hydrazine (GA
) 392.0 kcal/mol, value calculated at the DFT/6-311++G**
level)19 shows that the latter molecule is, in terms of
absolute acidity, 4.9 kcal/mol (or 3.6 pKa units) more
acidic than ammonia. This may be due to dominance of
the acidifying effect of the modestly electronegative
N-amino substituent over the increased lone-pair-lone-
pair repulsion acting in the opposite direction in the
anionic form of hydrazine (see also ref 20).
As mentioned in the Introduction, the pKa(DMSO) of the
extremely weak NH acids, ammonia and hydrazine,
cannot be determined experimentally due to autoprotoly-
sis of the solvent but the former has been estimated to
be 41 ( 1. The simplest possibility to estimate an
approximate pKa(DMSO) for hydrazine is now at hand in
eq 2, which leads to a value of 36.0 ( 1, indicating that,
like in the gas phase, hydrazine is by ca. 5 pKa units a
stronger acid than ammonia.
with parameters n ) 9, r ) 0.940, s ) 1.9
The roughly linear relationship (3) holds over a pKa(X-
NH2) interval of ca. 17 powers of 10 and shows that the
sensitivity toward substituent effects is slightly lower for
the hydrazines than that for the amines. On the basis of
eq 3, extrapolation to Bordwell’s value for ammonia (pKa
) 41) produces a very rough estimate (ca. 36.9) for the
pKa(DMSO) of hydrazine, again indicating that it is expected
to be somewhat more acidic than ammonia.
A third possibility to estimate a rough value of the pKa
of hydrazine in DMSO solution stems from earlier
findings by some of us11 that the experimental gas-phase
acidities in two reaction series (meta-substituted anilines
on one hand, and meta-substituted anilines and ammonia
on the other hand) depend linearly on their respective
pKa values in DMSO. The insertion of ∆Gacid ) 392.0 kcal/
mol19 for hydrazine into the corresponding equation11
leads to a third estimate (ca. 40 ( 1) for the pKa of
hydrazine in DMSO. The average of this value and that
obtained by extrapolation via eq 2 leads to the estimate
38.0 ( 1 for the pKa of hydrazine in DMSO, which would
make it roughly three pK units stronger acid than NH3.
In conclusion, it can be stated that for the protected
hydrazine compounds under study, the pKa(DMSO) values
distinctly reflect structural features and exhibit a total
span of about 21 units. Hydrazine derivatives with
typically three alkoxycarbonyl groups are weakest (Cat-
egory C), whereas such derivatives with one arenesulfo-
nyl instead of an alkoxycarbonyl group on the double-
protected nitrogen are of medium acidity (Category A).
N,N′-Diprotected hydrazines with trifyl, TFA, or arene-
sulfonyl-NH function are of highest acidity (Category B).
The novel pKa data rationalize previously reported work
with selected Category A and B reagents under Mit-
sunobu conditions as well as an unsuccessful experiment
with one such belonging to Category C.
Experimental Section
Another possibility to gain an approximate hypothetic
pKa(DMSO) for H2NNH2 is based on the observation that
Determination of pKa for Protected Hydrazines in
DMSO. The pKa determinations in DMSO were performed by
potentiometric titration in parallel with a solution of Bu4NOH
in a mixture of benzene and i-PrOH (4:1, v/v)7 and with a
solution of t-BuP4 superbase (Fluka) in a mixture of benzene
and DMSO (from 1:1 to 1:3 v/v, the concentration of titrant
was ca. 10-2 M and that of hydrazine ca. 10-3 M).7,11 In the
pKa range from ca. 9 to 18, the results in both titration series
were the same within the limits of 0.2-0.3 pKa units. It shows
(19) This work. No experimental value for the gas-phase acidity of
hydrazine is available. The ∆Gacid value given here (392.0 kcal/mol) is
the result of high-level quantum chemical calculations (DFT B3LYP
approach at the 6-311++G** level). At the more moderate 6-311+G**
level, this value becomes 391.7 kcal/mol (see Table 1A).
(20) Burk, P.; Koppel, I. A.; Rummel, A.; Trummal, A. J. Phys. Chem.
1995, 99, 1432.
5920 J. Org. Chem., Vol. 70, No. 15, 2005

Acidity of Di- and Triprotected Hydrazines
that the leveling effect of the water formed from the first
titrant is not significant, at least for pKa values up to 18. All
manipulations with the superbasic titrants were performed
in the Mbraun professional drybox. TFANHNHTFA (entry 12,
Part B) was synthesized according to a published procedure
and had the same data of analyses.21
1,1,2-Tri-tert-butyloxycarbonyl-2-methyl-hydrazine
(1,1,2-Boc3-2-Me-Hydrazine).(Useofatriflate.)BocNHNBoc24a,9e
(665 mg, 2.00 mmol) was dissolved in dry THF (4 mL) and
cooled to -78 °C under argon. A solution of BuLi (1.62 M in
hexane, 1.24 mL, 2.01 mmol) was then added dropwise with
rapid stirring over a period of 15 min and after a further 15
min cautiously treated dropwise with methyl triflate (394 mg,
2.40 mmol). After 1 h of being stirred at -78 °C and 2 h at
-30 ( 2 °C, the reaction mixture was diluted with Et2O (80
mL), washed in turn with 0.2 M citric acid, 1 M NaHCO3, and
brine (3 × each), and dried (MgSO4). The yield of pale yellow
oil was 695 mg (100%), essentially pure by 1H NMR and TLC
(Et2O/light petroleum 1:2). The analytical specimen was
obtained after chromatography on silica in the above solvent
mixture. 1H NMR: δ (major/minor conformer): 1.44/1.49 (2s,
9H), 1.513/1.505 (2s, 18H), 3.06/3.08 (2s, 3H). 13C NMR: δ
(major/minor conformer): 27.9, 28.13/28.17, 35.6/37.0, 81.0/
81.2, 83.2, 150.0/150.3, 153.86/153.81.
1-Benzyl-1-benzyloxycarbonyl-2-tert-butyloxycarbonyl-
2-(4-toluenesulfonyl)-hydrazine (1-Bn-2-Boc-2-Ts-1-Z-Hy-
drazine). (Typical alkylation procedure using an alcohol under
Mitsunobu conditions.) A solution of ZNHNTsBoc9a (1.26 g,
3.00 mmol) and benzyl alcohol (357 mg, 3.30 mmol) in dry THF
(10 mL) was chilled to -10 °C under argon, and recrystallized
triphenylphosphine (946 mg, 3.60 mmol) was introduced slowly
with rapid stirring. The resulting clear solution was then
treated dropwise, under thorough mixing, with diethyl azodi-
carboxylate (678 mg, 3.90 mmol) in THF (3 mL) over a period
of 30 min and then allowed to react for 1 h, whereafter the
cooling bath was removed. After 16 h, TLC (CH2Cl2/Et2O 9:1)
indicated complete reaction, and most of the solvent was
stripped off at reduced pressure. The yellowish oily residue
was dissolved in CH2Cl2 and chromatographed on silica
(chromatographic system as above), providing 1.52 g (99%) of
the pure title compound, a white, microcrystalline solid, mp
89.5-90 °C (from Et2O; ∼5 mL/g; -20 °C), in all respect
identical with an authentic sample.9a
Acknowledgment. The work was supported in part
by Helge Ax:son Johnson’s stiftelse and the Estonian
Science Foundation (Grants No. 5226 and 5255). U.R.
is grateful to the Royal Swedish Academy of Sciences
and the Estonian Academy of Sciences for a scientific
exchange scholarship.
1-Benzyl-2-tert-butyloxycarbonyl-1-(4-toluenesulfonyl)-
hydrazine (1-Bn-2-Boc-1-Ts-Hydrazine). (Example illus-
trating benzylation on sulfonyl nitrogen of a diprotected
reagent under Mitsunobu conditions.) This synthesis was
described elsewhere.4c
Supporting Information Available: Synthetic proce-
dures for novel substances including spectral data and results
of DFT calculations of acidities of hydrazines. This material
is available free of charge via the Internet at http://pubs.acs.org.
(21) Groth, R. H. J. Org. Chem. 1960, 25, 102.
JO050680U
J. Org. Chem, Vol. 70, No. 15, 2005 5921