Reductive demercuration in deprotection of trityl thioethers, trityl amines, and trityl ethers

Article abstract of DOI:10.1021/jo0156971

A room-temperature deprotection method of trityl amines, -ethers, and -thioethers is presented, based on coupling of metal acid catalysis (HgX₂, with X^- = Cl^- or OAc^-) and sodium borohydride reduction. The results of its application to monotritylated compounds (ethanethiol, ethanol, and piperidine) and to mono- and ditritylated 1,2-bifunctional compounds (mercaptoethanol, aminoethanethiol, and ethanolamine) are compared with those obtained with early methods based on the use of strong Bronsted acids (pure TFA and MeCN solutions of HCl). Trityl thioethers of simple thiols and amino and hydroxy thiols are promptly cleaved by reductive detritylation, and one-pot procedures can be employed to produce free thiols. In contrast, dilution with water of these same compounds in solutions of strong Bronsted acids leaves them unaffected. O-Tr and N-Tr bonds are broken by this latter treatment. However, trityl ethers are rapidly cleaved by even dilute HCl solutions, while cleaving of trityl amines is modulated by HCl concentration. Addition of NaBH₄ to solutions of monofunctional trityl ethers in HgCl₂/MeCN leads to complete deprotection. Monofunctional trityl amines are partially deprotected only if the complexation reaction is allowed to reach equilibrium. Combination of H⁺- with HgX⁺-catalyzed detritylation methods allows selective deprotection of pertritylated amino and hydroxy thiols. The results appear to be due to the strong difference in the affinity of the donor atoms present in the pertritylated substrates for H⁺ and HgX⁺. Catalysis based on Bronsted acids leads to cleaving of the N- and O-trityl bonds with recovering of the S-trityl group; that based on mercury salts allows recovering of N- and O-trityl groups with deprotection of the -SH function. Selectivity in deprotection of pertritylated amino alcohols seems to be severely hampered by similarity in the affinity of N- and O-atoms for H⁺ and HgX⁺, and, taking advantage of the lower HgX⁺-complexation rate of the N-trityl with respect to the O-trityl group, only preservation of the N-trityl bond has been achieved.

Full text of DOI:10.1021/jo0156971

J . Org. Chem. 2001, 66, 7615-7625
7615
Red u ctive Dem er cu r a tion in Dep r otection of Tr ityl Th ioeth er s,
Tr ityl Am in es, a n d Tr ityl Eth er s
M. Maltese
Dipartimento di Chimica, Universita` “La Sapienza”, p.le A. Moro 5, Box 34 Roma 62, Italy
maurizio.maltese@uniroma1.it
Received April 19, 2001
A room-temperature deprotection method of trityl amines, -ethers, and -thioethers is presented,
based on coupling of metal acid catalysis (HgX2, with X- ) Cl- or OAc-) and sodium borohydride
reduction. The results of its application to monotritylated compounds (ethanethiol, ethanol, and
piperidine) and to mono- and ditritylated 1,2-bifunctional compounds (mercaptoethanol, aminoet-
hanethiol, and ethanolamine) are compared with those obtained with early methods based on the
use of strong Brønsted acids (pure TFA and MeCN solutions of HCl). Trityl thioethers of simple
thiols and amino and hydroxy thiols are promptly cleaved by reductive detritylation, and one-pot
procedures can be employed to produce free thiols. In contrast, dilution with water of these same
compounds in solutions of strong Brønsted acids leaves them unaffected. O-Tr and N-Tr bonds
are broken by this latter treatment. However, trityl ethers are rapidly cleaved by even dilute HCl
solutions, while cleaving of trityl amines is modulated by HCl concentration. Addition of NaBH4 to
solutions of monofunctional trityl ethers in HgCl2/MeCN leads to complete deprotection. Mono-
functional trityl amines are partially deprotected only if the complexation reaction is allowed to
reach equilibrium. Combination of H+- with HgX+-catalyzed detritylation methods allows selective
deprotection of pertritylated amino and hydroxy thiols. The results appear to be due to the strong
difference in the affinity of the donor atoms present in the pertritylated substrates for H+ and
HgX+. Catalysis based on Brønsted acids leads to cleaving of the N- and O-trityl bonds with
recovering of the S-trityl group; that based on mercury salts allows recovering of N- and O-trityl
groups with deprotection of the -SH function. Selectivity in deprotection of pertritylated amino
alcohols seems to be severely hampered by similarity in the affinity of N- and O-atoms for H+ and
HgX+, and, taking advantage of the lower HgX+-complexation rate of the N-trityl with respect to
the O-trityl group, only preservation of the N-trityl bond has been achieved.
In tr od u ction
reaction mixture after deprotection. In analogous fashion,
acid-catalyzed N- and O-detritylations are reported to be
Triphenylmethyl is a common protecting group for
hydroxyls and primary and secondary amino and thiol
groups, and acidolysis represents the most practical
method to remove it from the protected groups at the end
of any desired reaction of the molecule under study. In
principle, selective deprotection of pertritylated com-
pounds, with at least two of the above listed functional
groups protected, should be gained through controlled
acidolysis, factors kinetically affecting the process being
the strength of the Brønsted acid employed, its concen-
tration, temperature, etc. However, data coming from the
literature on deprotection times and temperatures of
trityl thioethers, amines, and ethers1 appear too much
spread to put some confidence in selectively detritylating
these groups. For example, S-detritylations with TFA
have been performed at temperatures ranging from room
to reflux temperature and required from 15 to 30 min.2,3
Remarkably, the use of this as well as other strong protic
acids has been found ineffective in many cases.4 Further
differences can be found in methods of working up the
achieved in a variety of conditions.1
On the other hand, it is well-known that heavy metal
ions displace trityl groups from trityl thioethers, both in
protic and aprotic solvents, and in the former case,
irrespective of the solvent acid strength.1 With respect
to acidolysis through a Brønsted acid, deprotection of
trityl thioethers performed with the aid of metal ions
introduces a further step, represented by the displace-
ment of the metal from the mercaptide, usually per-
formed by treatment with competing thiols or with acids,
as HCl or H2S, having anions giving rise to insoluble salts
with the metal ion.1,4 Furthermore, the use of metal ions
in deprotecting trityl amines and trityl ethers appears
to be much less investigated.
An early goal of this work was the characterization of
the deprotection process of trityl thioethers, primarily
with metal acids in aprotic solvents. Then, it was found
that trityl amines and trityl ethers are detritylated by
TFA at room temperature at a rate definitely higher than
trityl thioethers, thus pointing to the concrete existence
of methods of selective detritylation based on the kinetic
control of the H+-catalyzed reaction. In a further step,
the encouraging results obtained in the recovery of thiols
* To whom correspondence should be addressed. Fax: +39(0)-
6490324.
(1) Greene, T. W.; Wuts, P. G. M. Protective groups in organic
synthesis; Wiley: New York, 1991; Chapters 2, 6, and 7.
(2) Zervas, L.; Photaki, I. J . Am Chem. Soc. 1962, 84, 3887.
(3) Photaki, I.; Taylor-Papadimitriou, J .; Sakarellos, C.; Mazarakis,
P.; Zervas, L. J . Chem. Soc. C 1970, 2683.
(4) Carroll, F. I.; Dickson, H. M.; Wall, M. E. J . Org. Chem. 1965,
30, 33.
10.1021/jo0156971 CCC: $20.00 © 2001 American Chemical Society
Published on Web 10/12/2001

7616 J . Org. Chem., Vol. 66, No. 23, 2001
Maltese
from their mercaptides by treatment with sodium boro-
hydride prompted an extension of the method to trityl
amines and trityl ethers.
of each species taking part in eq 1, in particular with
NaBH4, has been accurately investigated.
In TFA, trityl derivatives of monofunctional amines,
thiols, and alcohols give rise to yellow solutions whose
rapid dilution with water at room temperature leads to
prompt cleavage only of N-Tr and O-Tr bonds (hereto-
fore, this procedure will be concisely indicated as TFA,
H2O; symbolism for this and other detritylation proce-
dures is described in the Experimental Section). In
contrast, trityl thioethers coming from simple thiols and
amino and hydroxy thiols remain unaffected even after
prolonged room-temperature treatment with TFA. AcOH
does not deprotect trityl thioethers and, as an effect of
its weakness, does not affect, as well, trityl amines and
trityl ethers: dilution with water of AcOH solutions of
these substrates provided only recovered starting mate-
rial. Thus, even strong protic acids do not cleave S-Tr
bonds, while only strong acids break N-Tr and O-Tr
bonds. Remarkably, detritylation with TFA is one of the
recommended methods for trityl thioethers.1
For the sake of comparison with the use of MeCN
solutions of HgX2 (X ) Cl-, AcO-), described below,
detritylation has been attempted also with HCl solutions
in MeCN (procedure HCl/MeCN, H2O). Trityl thioethers
(Et-STr, ME-STr) were moderately detritylated only by
strongly concentrated solutions of this acid, but with
formation of secondary products. Trityl ethers (Et-OTr)
were totally cleaved while cleaving of trityl amines (Pip-
Tr and EA-NTr) is modulated from partial to complete
by HCl concentration
A TFA or AcOH solution of a mercuric salt HgX2
cleaves the S-Tr bond, and dilution with water gives rise
to formation of the mercury mercaptide and TrOH. The
acid strength does not affect the result, according to early
successful quantitative S-detritylations with Hg(OAc)2 in
EtOH4 and in MeOH.2 Trityl amines or ethers are fully
deprotected by HgX2/TFA, H2O, but this result seems to
be due to the strength of TFA alone. The addition of the
metal acid to AcOH (procedure HgX2/AcOH, H2O) gives
rise to negligible traces of N- and O-detritylation. The
introduction of the investigated substrates in a HgX2/
MeCN solution, followed by addition of water (HgX2/
MeCN, H2O), gives rise to S-detritylation but leaves
unaffected N-Tr and O-Tr bonds.
On passing to the use of NaBH4 as nucleophile, it is
clear that the final product of an its reaction on a given
substrate does not necessarily depend on the substrate
structure (known or merely hypothesized) in its solid
state. Nevertheless, the investigation of reactions of this
reductant with isolated and well characterized sub-
stances, taking part to equilibrium (1) or related with it,
can throw light on the results reported in this paper. So,
simple trityl derivatives are directly deprotected by
reduction with NaBH4/MeCN: TrCl and TrOAc have
been rapidly converted into TrH. The same product has
been reported as prevalent when TrPF6 is reduced with
NaBH4 in THF.6 This result has to be ascribed to Cl- and
AcO- being good leaving groups. Accordingly, triphenyl-
carbinol and all the investigated substrates are not
affected by this same treatment.
This work reports the results of detritylation experi-
ments on the substrates listed in the Experimental
Section.5 The idea was to compare the efficiency of early
procedures with the detritylation method here investi-
gated first on tritylated monofunctional compounds (eth-
ylmercaptan, ethanol, and piperidine), then on trityl
derivatives of compounds bearing two different functions
(mercaptoethanol, ethanolamine, and mercaptoethyl-
amine) at various levels of protection (for example, S- and
O-tritylmercaptoethanol and O,S-ditritylmercaptoetha-
nol). The behavior under detritylation of the bis-tritylated
compounds has given information about selective depro-
tection, while that of the monotritylated substrates has
shed light on the influence of the free group on the
detritylation of the other (Tr-migration and so on).
As will be shown below, the results allow useful rules
for selective detritylation to be drawn. The work herein
reported encourages a broader investigation on selective
protection or deprotection of multifunctional substrates
under metal-acid catalysis, and experiments in this area
are currently in progress in this laboratory.
Resu lts
Detritylation requires the action of a nucleophile on
the following general equilibrium, where D is the donor
atom to be deprotected and A+ is the Lewis acid employed
as catalyst:
(reference to formalism introduced with eq 1 will be done
throughout the following text). The experimental results
indicate as fast the reactions involved in eq 1 when A+
) H+, while HgCl+ complexation by trityl amines appears
to be slow (see below). Competition between the reactions
of a given nucleophile with many substances taking part
to equilibrium (1) is to be expected, and which route
prevails depends on the equilibrium concentrations of
A+X-, Tr+, TrX, of the adduct R-D-A and of the inter-
mediate complex, on the nature of the employed nucleo-
phile, and on the relative reaction rate of each of the
species reacting with the given nucleophile. For a deeper
insight in the results obtained in this work, the reactivity
(5) Notation: ME ) mercaptoethanol; MEA ) aminoethanethiol;
EA ) ethanolamine; TrCl ) triphenylmethyl chloride; TrOH ) tri-
phenylcarbinol; TrOAc ) triphenylmethyl acetate; TrH ) triphenyl-
methane; Et-STr ) trityl-ethanethiol; Et-OTr ) trityl-ethanol; Pip-Tr
) tritylpiperidine; ME-STr ) S-tritylmercaptoethanol; ME-OSTr )
O,S-ditritylmercaptoethanol; ME-OTr ) O-tritylmercaptoethanol; MEA-
STr: S-tritylaminoethanethiol; MEA-NSTr
) N,S-ditritylamino-
The hydrochlorides of EA-NTr and Pip-Tr, as repre-
sentative isolated intermediate complexes of trityl amines
of the type (R-D+(H)-Tr)Cl-, react with NaBH4/MeCN or
THF giving rise to hydrogen and to the parent trityl
ethanethiol; MEA-NTr ) N-tritylaminoethanethiol; EA-NTr ) N-
tritylethanolamine; EA-NOTr ) N,O-ditritylethanolamine; EA-OTr )
O-tritylethanolamine; CYA-2Tr ) N,N′-ditrityl-2-aminoethyl disulfide
(N,N′-ditritylcystamine); MED ) 2-hydroxyethyl disulfide; MED-2Tr
) O,O′-ditrityl-2-hydroxyethyl disulfide; ME-SHgX, ME-OTrSHgX,
MEA-SHgX and MEA-NTrSHgX (with X ) Cl, OAc or OTFA) indicate
mercaptides of ME, ME-OTr, MEA, and MEA-NTr, respectively.
(6) Olah, G. A.; Svoboda, J . J . J . Am. Chem. Soc. 1973, 95, 3794.

Reductive Demercuration in Deprotection of Trityl Thioethers
J . Org. Chem., Vol. 66, No. 23, 2001 7617
F igu r e 1. Deprotection pathways of tritylated derivatives of a representative 1,2-amino thiol (mercaptoethylamine).
amine, even in the presence of an excess of HCl. Dropping
of a HCl/MeCN solution of a trityl ether into a NaBH4/
MeCN suspension leads to complete detritylation, but
this result can be confidently referred to the catalytic
effect of HCl alone. Trityl thioethers remain essentially
unaffected by this same treatment. Remarkably, the
procedure HCl/MeCN, NaBH4 leaves unaffected the
N-Tr bonds, while it leads to complete O-detritylation.
On addition to HgX2/MeCN, all the investigated trityl
derivatives give rise to coordination of mercury, and TLC
can help in having an insight in relative amounts of
species taking part to equilibrium (1) (see the Experi-
mental Section). In many instances, complexation leads
to heterolytic release of Tr+ (as indicated by the more or
less intense yellow color of the solution7,8). More gener-
ally, the resulting solutions are colorless. On the grounds
of results coming from TLC controls, differences in rates
of HgCl+-complexing reactions have been revealed. It has
been found that yellow solutions of trityl thioethers and
colorless solutions of trityl ethers in HgCl2/MeCN go
rapidly to equilibrium, while trityl amines require hours
to give a constant chromatogram. Reaction with NaBH4
of freshly prepared solutions of Pip-Tr does not lead to
detritylation. Solutions of Pip-Tr, left aside for 24 h, give
rise to precipitation of a solid in which the presence of
the moiety (Pip+(HgCl)-Tr)Cl- can be confidently as-
sumed (see the Experimental Section). In the solution
in equilibrium with this solid the free substrate is
practically absent. Reduction of both the suspension and
of the isolated solid leads to only partial detritylation. A
different result has been obtained for EA-NTr. In this
case, complication is introduced by metal acid catalyzed
Tr-migration: after 48 h from formation of the mixture,
TLC shows disappearance of substrate, but appearance
of EA-NOTr. However, reduction with NaBH4 of this
mixture has given only EA-NTr as final product, in
agreement with the expectation that N-Tr groups are
only slightly affected by this method.
As far as the reactivity of adducts of the type R-D-HgCl
with NaBH4 is concerned, ME-SHgCl, easily synthesized
from ME in excess HgCl2, reduced with NaBH4 in MeCN,
has given free ME and Hg0. In contrast, attempts to
prepare the adduct Pip-HgCl failed. However, the solid
obtained by adding Pip to excess HgCl2 in MeCN, having
a composition corresponding to (Pip+(H)-HgCl)Cl-, is
reduced by NaBH4 to Pip and Hg0. By comparison with
the behavior of the hydrochloride (Pip+(H)-Tr)Cl-, re-
ported above, it can be expected that, in the reaction with
NaBH4, protons would be first removed from (Pip+(H)-
HgCl)Cl-, with production of hydrogen and of the adduct
Pip-HgCl. Then, the attack on this compound should take
place, with production of Hg0 and Pip, as experimentally
observed.
The addition of NaBH4 to simple trityl thioethers and
ethers (Et-STr and Et-OTr) in HgCl2/MeCN (procedure
HgCl2/MeCN, NaBH4, see the Experimental Section),
completed with introduction of water in the reaction
mixture, leads to quantitative deprotection. In contrast
with this indication, the O-Tr group, in the presence of
S-Tr or -SH moieties on the same molecule, remains
unaffected by this same treatment: ME-OSTr is only
S-detritylated by NaBH4, with preservation of the O-Tr
bond.
Figures 1-3 summarize the results of the application
of procedures giving rise to effective detritylation in
mono- and ditritylated 1,2-aminothiols, -hydroxythiols,
and -amino alcohols, respectively. The results point to
practical routes to selective detritylation of substrates
bearing two different functions. So, the treatment of a
(7) Bentley, A.; Evans, A. G.; Halpern, J . Trans. Faraday Soc. 1951,
47, 711.
(8) All the TFA solutions of the substrates under investigation and
part of them in HgCl2/MeCN have a bright yellow colour, due to the
presence of the triphenylcarbenium ion. The solutions of these same
substances in AcOH or Hg(OAc)2/MeCN are colorless.

7618 J . Org. Chem., Vol. 66, No. 23, 2001
Maltese
F igu r e 2. Deprotection pathways of tritylated derivatives of a representative 1,2-hydroxy thiol (mercaptoethanol).
NaBH4 treatment leads to the fully deprotected amino-
thiol.
In analogous fashion (see Figure 2), a pertritylated
hydroxy thiol such as ME-OSTr, when treated with HCl/
MeCN, H2O, gives the S-tritylated hydroxy thiol, ME-
STr. The low solubility of this substrate in TFA does not
recommend the TFA, H2O procedure for its O-detrityl-
ation. The thiol group of ME-STr can be deprotected by
the one-pot procedure Hg(OAc)2/MeCN, NaBH4, or, step
by step, through the isolation of the mercaptide ME-
SHgX and further reduction with NaBH4/MeCN. In con-
trast, O-tritylation can be preserved if the pertritylated
substrate is first treated with Hg(OAc)2/AcOH or MeCN,
NaBH4, and the resulting O-tritylated derivative is fully
deprotected to the original hydroxy-thiol by treatment
with HgX2/TFA, H2O, NaBH4. The mercury mercaptide
ME-SHgCl can be obtained from the pertritylated sub-
strate by HgCl2/HCl/MeCN, H2O, while ME-OTrSHgOAc
has been isolated by treatment of the same substrate
with Hg(OAc)2/AcOH, H2O. Complete detritylation of ME-
OSTr can be achieved by HgX2/TFA, H2O, NaBH4.
Pertritylated amino alcohols appear to be less versatile
substrates for selective detritylation. The N-Tr bond is
preserved under treatment by HgCl2/MeCN, NaBH4, and
the resulting N-protected amino alcohol is fully detri-
tylated by TFA, H2O. This same product, EA-NTr, has
been obtained when EA-NOTr is treated with HCl/MeCN,
H2O. Complete deprotection of the pertritylated substrate
is obtained by TFA, H2O, and one has the same result
when the substrate is EA-OTr. However, attempts to
selectively detritylate -NTr in the presence of -OTr led
always to mixtures of the N- and O-monotritylated
derivatives, pointing to a preferential detritylation of the
-OTr group.
F igu r e 3. Deprotection pathways of tritylated derivatives of
a representative 1,2-amino alcohol (ethanolamine).
pertritylated amino thiol, say MEA-NSTr, with TFA, H2O
(Figure 1) leads rapidly to the N-detritylated analogue.
Further detritylation of this latter substrate can be
achieved by treatment with Hg(OAc)2/AcOH, H2O or
HgCl2/MeCN, H2O, followed by reduction with NaBH4:
the intermediate mercaptide MEA-SHgX can be isolated,
then reduced by NaBH4/MeCN, or the reaction can be
realized in a one-pot procedure. The same pertritylated
amino thiol can be S-deprotected by treatment with
Hg(OAc)2/AcOH or HgX2/MeCN, followed by reduction
with NaBH4 of the isolated mercaptide, or in the one-
pot procedure. The resulting NTr-aminothiol cannot be
fully deprotected to the free aminothiol by TFA, H2O,
since acid-catalyzed Tr-migration to the S-atom occurs.
Finally, the HgX2/TFA, H2O treatment on the per-
tritylated aminothiol MEA-NSTr gives the mercaptide
MEA-SHgX, while the one-pot procedure HgX2/TFA, H2O,
Exp er im en ta l Section
Melting points were determined with a Kofler hot-stage
1
apparatus and are uncorrected. IR and H NMR spectra of the

Reductive Demercuration in Deprotection of Trityl Thioethers
J . Org. Chem., Vol. 66, No. 23, 2001 7619
synthesized substrates have been found in agreement with the
expectations and not reported. Thin-layer chromatography was
carried out on Merck HF254 silica gel, using three systems as
eluents: A, CHCl3; B, CHCl3/n-Heptane 1:1; C, CHCl3/MeCN
4:1. Besides UV light, the spots have been characterized with
ninhydrin (for amines and trityl amines), TFA vapors (which
develop a bright yellow color in the presence of trityl-residues
in the molecule), NaBH4 in NaOH 3 M (which develops a black
color in the presence of reducible mercury compounds), and
Ni(OAc)2/MeOH (which sprayed onto TLC plates, develops
brown colors in the presence of thiol groups). For microanaly-
ses, I am indebted to Dr. P. Galli of this same department,
whose collaboration is gratefully acknowledged. All the com-
mercial chemicals employed were of reagent grade, used,
unless otherwise stated, without further purification and
obtained from Merck, C. Erba, and Aldrich. Water was
removed from commercial MeCN and ethanolamine by azeo-
tropic distillation with benzene and fractionation. Diethyl
ether was distilled from LiAlH4 to remove moisture. THF was
treated with NaOH pellets to remove peroxides and dried as
for diethyl ether.
Su bstr a tes. 1. Tr ityleth a n eth iol (Et-STr ). Ethanethiol
(1.00 g, 16 mmol) was rapidly added under stirring to a
suspension of 4.50 g (16 mmol) of triphenylmethyl chloride and
1.8 g (18 mmol) of TEA in 20 mL of MeCN. The mixture, left
aside for 1 h, was poured into 200 mL of water, the resulting
suspension was filtered off and the solid washed with water.
The crude product (4.58 g, th. 4.91 g, 92.26%), slightly impure
for TrOH, has been purified by crystallization from EtOH, mp
126-27 °C (lit.9 mp 125 °C).
MEA-NTr. In an experiment, 0.8 mmol of TrOH was recovered.
TLC analysis of the final product, along with ME-OTr,
indicated the presence of ME-OSTr and ME-STr, coming from
O- to S-trityl migration in ME-OTr. Since the reactivity of ME-
OTr has prevented any further purification, its recovery has
been accomplished by oxidation to its disulfide, MED-2Tr. The
crude product was dissolved in Et and oxidized with I2/Et in
the presence of stoichiometric TEA (0.9 mmol). The final
mixture was extracted with water, and the Et solution was
dried and evaporated. The crude product was crystallized from
MeCN, obtaining 0.18 g of MED-2Tr (th. 0.29, 64.07%), mp
93-94 °C.
Anal. Calcd for C42H38O2S2: C, 78.96; H, 5.99; S, 10.04.
Found: C, 78.58; H, 5.81; S, 9.87.
An authentic sample of MED-2Tr was synthesized through
I2 oxidation of an Et solution of ME in the presence of solid
NaHCO3. The final mixture was filtered off, the solid was
thoroughly washed with Et, and the Et solution was evapo-
rated under vacuum. An oil was obtained (MED) which,
dissolved in Py, along with a small excess of TrCl, was left
aside for 3 days at rt and then poured into 150 mL of water.
The suspension was slightly acidified with HCl and the solid
filtered off and washed with water.
5. S-Tr ityla m in oeth a n eth iol (MEA-STr ). Aminoethane-
thiol hydrochloride (2.00 g, 17.6 mmol) was dissolved in 5 mL
of TFA. Triphenylcarbinol (4.58 g, 17.6 mmol) was added to
the solution portionwise, under stirring at room temperature,
until the solution became clear. The reaction mixture, a dense
deeply red liquid, was left aside for 1 h and then poured in
200 mL of water under vigorous stirring. The suspension of
the white solid was alkalanized with TEA and filtered and the
solid washed with water alkaline for TEA. After drying, 5.60
g of the crude solid, slightly impure for TrOH, was obtained
(th. 5.62 g). Purification was achieved by dissolving the crude
product in HCl/MeCN and rapidly diluting 1:10 with com-
mercial Et the clear solution, obtaining a white fibrous solid,
Anal. Calcd for C21H20S: C, 82.85; H, 6.62; S, 10.53. Found:
C 82.76, H 6.71, S 10.44.
2. S-Tr itylm er ca p toeth a n ol (ME-STr ). Solid triphenyl-
carbinol (1.67 g, 6.4 mmol) was rapidly added, under stirring
at room temperature, to a solution of 0.50 g of mercaptoethanol
(6.4 mmol) and 0.63 g of HCl 37% in 8 mL of MeCN. The
crystalline solid, precipitated from the clear solution, was
stirred for 10 min, filtered, and washed with MeCN, obtaining
1.3 g of ME-STr chromatographically pure (th. 2.05 g, 63.41%).
An analytical sample, obtained by further recrystallization
from the same solvent, had mp 116-117 °C.
the monohydrated hydrochloride (Anal. Calcd for C21H24
-
ClNOS: C, 67.45; H, 6.47; N, 3.75; S, 8.57. Found: C, 67.63;
H, 6.63; N, 3.73; S, 8.48). The base can be obtained from this
solid, after filtration and washing with Et, by stirring an its
suspension in Et/concentrated KOH. When both phases were
clear, the organic layer was isolated and dried. The solid
obtained by evaporation of the Et solution was chromato-
graphically pure. An analytical sample was obtained by further
crystallization from MeCN, mp 91-93 °C (lit.4 mp 90-93 °C).
Anal. Calcd for C21H21NS: C, 78.95; H, 6.63; N, 4.38; S,
10.04. Found: C, 78.85; H, 6.68; N, 4.41; S, 9.92.
Alternatively, MEA-STr has been prepared by selective
detritylation of MEA-NSTr with TFA: 1.00 g (1.8 mmol) of
the pertritylated MEA was dissolved in 5 mL of TFA and left
aside for 3 min. Then, water was added (10:1) and the resulting
suspension was alkalanized with concentrated KOH and
filtered off. The solid, an equimolar mixture of MEA-STr and
TrOH, was purified as before through the formation of the
hydrochloride. Melting point and analytical data have been
found consistent with those reported above.
Anal. Calcd for C21H20OS: C, 78.71; H, 6.29; S, 10.00.
Found: C, 78.62; H, 6.32; S, 9.92.
Alternatively, ME-STr has been obtained by detritylating
ME-OSTr with HCl/MeCN: 1.00 g (1.8 mmol) of ME-OSTr was
dissolved in 6 mL of HCl/MeCN 0.1 M and left aside at room
temperature for 5 min, then the mixture was poured in water
(120 mL) and the suspension neutralized with NaHCO3,
filtered, and the solid washed with water. The desired product
was isolated from TrOH by washing the solid with cold MeOH
and filtering the suspension. The addition of water 5:1 to the
filtrate has given a crude solid which, after filtration and
crystallization with MeCN produced a chromatographically
pure ME-STr with a yield of 67%.
3. O,S-Ditr itylm er ca p toeth a n ol (ME-OSTr ). A 0.30 g
portion of mercaptoethanol (3.8 mmol) was added to a solution
of 2.30 g (8.8 mmol) of TrOH in 5 mL of TFA, and the mixture
was kept under stirring for 3 h, poured under vigorous stirring
in 100 mL of water, and neutralized with concentrated KOH.
The mother liquor was decanted and the sticky paste washed
with water, dissolved in 5 mL of hot MeCN, and left aside for
12 h. The crystalline solid, filtered and washed with MeCN
(1.80 g, th. 2.49 g, 72.29%), contained only traces of ME-STr.
The analytical sample was obtained by two crystallizations
from MeCN, mp 173-175 °C.
6. N,S-Ditr ityla m in oeth a n eth iol (MEA-NSTr ). A 5.34 g
(53 mmol) portion of TEA was added to a solution of 2.00 g
(18 mmol) of aminoethanethiol hydrochloride in 10 mL of
DMF. To the obtained suspension, kept under stirring, 10.32
g (37 mmol) of TrCl was added portionwise. Then, the mixture
was left aside for 4 h. After the mixture was poured into 200
mL of water, the liquid was decanted and the sticky paste
washed with water and triturated with EtOH, giving rise to a
white solid, whose weight, after filtration, washing with EtOH,
and drying, was 21.88 g (th. 29.78 g, 73.49%). The analytical
sample was obtained after two crystallizations from MeCN,
mp 161-63 °C.
Anal. Calcd for C40H34OS: C, 85.37; H, 6.09; S, 5.70.
Found: C, 85.21; H, 6.15; S, 5.78.
4. O-Tr itylm er ca p toeth a n ol (ME-OTr ). A 0.50 g portion
of ME-OSTr (0.9 mmol) was added to a solution of 1.0 g of
Hg(OAc)2 (3.1 mmol) in 6 mL of AcOH and left under stirring
to complete dissolution. Then, 45 mL of water was added, and
the resulting suspension was treated as described below for
Anal. Calcd for C40H35NS: C, 85.52; H, 6.28; N, 2.49; S, 5.71.
Found: C, 85.23; H, 6.37; N, 2.60; S, 5.80.
MEA-NSTr was prepared also by addition of triphenyl-
methyl chloride (0.87 g, 3.1 mmol) to a solution of 1.00 g (3.1
mmol) of MEA-STr and 0.32 g of TEA (3.2 mmol) in 4 mL of
DMF. The mixture was allowed to react under stirring at room
temperature, and then it was poured in 100 mL of water and
(9) Fischer, P. J . Prakt. Chem. 1910, 82 (2), 523.

7620 J . Org. Chem., Vol. 66, No. 23, 2001
Maltese
TEA was added up to pH 10. The solid was filtered, washed
with water/TEA, dried, and crystallized from DMF (1.00 g, th
1.76 g, 57.0%).
Py. The mixture was stirred at room temperature for 3 days
and then poured under vigorous stirring in 250 mL of water.
The resulting suspension was filtered and the solid washed
with water, dried under vacuum, and dissolved in ether. This
solution, transferred in a separatory funnel, was treated
rapidly with HCl/H2O 3 M, and the obtained solid was filtered
off, washed with water, thoroughly dried, and washed with
Et. EA-OTr was regenerated by stirring the suspension of the
solid in Et/concentrated KOH. The clear organic phase was
washed with water, dried and evaporated, obtaining 3 g of
chromatographically pure solid (th. 6.52 g, 46.0%) with mp 89-
92 °C, employed as analytical sample.
7. N-Tr ityla m in oeth a n eth iol (MEA-NTr ). MEA-NSTr
(0.63 g, 1.1 mmol) and 1,80 g of Hg(OAc)2 (5.6 mmol) were
dissolved in 20 mL of AcOH, the clear solution was stirred for
2 min, and then water was added to a volume of 120 mL. The
resulting suspension of MEA-NTrSHgOAc and TrOH was
neutralized with concentrated KOH and filtered off. The solid,
washed with water and dried under vacuum, was suspended
in Et until triphenylcarbinol disappeared and then filtered off.
The Et solution, passed through silica gel and evaporated, gave
1.0 mmol of chromatographically pure TrOH. The residue was
suspended in 20 mL of MeCN, and solid NaBH4 was added
under stirring to this suspension. The reaction was considered
to be complete 15 min after the start. After the addition of 10
mL of Et, water was introduced cautiously and the suspension
left under stirring until gas evolution was almost completely
ceased. Then, borine complexes were destroyed with HCl 1:1,
the pH was adjusted to 7-8 with KOH and AcONa, and the
mixture was filtered off. The two-phase filtrate was transferred
in a separatory funnel and the organic layer was isolated, dried
on Na2SO4 and rapidly evaporated under vacuum. TLC
analysis on the crude solid, along with MEA-NTr as the
principal product, pointed to the presence of MEA-NSTr and
MEA-STr, coming from N to S trityl migration in MEA-NTr.
Owing to the reactivity of this substrate, its recovering has
been accomplished by oxidation to its disulfide, CYA-2Tr. The
crude residue was dissolved in Et and titrated with I2/Et. Soon
in the course of titration, a yellow solid precipitated (the
hydroiodide of CYA-2Tr), which, filtered and washed with Et,
suspended in Et and kept under stirring in the presence of
concentrated KOH has given rise to an ethereal solution of
CYA-2Tr. After this organic solution was dried, CYA-2Tr was
recovered chromatographically pure by solvent evaporation
(0.23 g, th. 0.35 g, 65.8%).
The solid was found identical with an authentic sample
prepared from cystamine hydrochloride by treatment with
TrCl/DMF in the presence of TEA. CYA-2Tr, crystallized from
MeCN, had mp 141-142 °C, and an analytical sample has
given the following data. Anal. Calcd for C42H40N2S2: C, 79.20;
H, 6.33; N, 4.40; S, 10.07. Found: C, 79.08; H, 6.43; N, 4.59;
S, 9.88.
In an attempt to prepare MEA-NTr through a one-pot
procedure, NaBH4 was directly added to the neutralized water
suspension of MEA-NTrSHgOAc. However, this procedure led
to negligible reduction of the mercaptide, which has been
recovered as a black gummy solid at the end of reaction. For
one-pot S-detritylations of MEA-NSTr, see below in the
detritylation experiments section.
Anal. Calcd for C21H21NO: C, 83.16; H, 6.98; N, 4.62.
Found: C, 82.98; H, 7.29; N, 4.52.
11. Tr ityleth a n ol (Et-OTr ). A mixture of excess anhydrous
EtOH, 2.00 g of TrCl (7.2 mmol), and 0.90 g of TEA (8.9 mmol)
was left aside for 24 h. Then, the liquid was distilled under
vacuum. The addition of 5 mL of MeOH to the sticky solid
residue gave rise to the formation of 0.8 g of chromatographi-
cally pure crystals (th. 2.07 g, 38.7%). The analytical sample
was obtained by further crystallization from EtOH, mp 81-
83 °C (lit.4,10 mp 79-82 °C and 84-85 °C, respectively).
Anal. Calcd for C21H20O: C, 87.46; H, 6.99. Found: C, 87.46;
H, 7.21.
12. Tr itylp ip er id in e (P ip -Tr ). Piperidine (0.70 g, 8.0
mmol) was added to a hot solution of 1.00 g of TrCl (3.6 mmol)
in 3 mL of MeCN, and the mixture was left aside for 30 min
and then poured under stirring in 150 mL of water. The
resulting suspension was extracted with Et. The organic
extract, after drying over Na2SO4, was evaporated and the
residue crystallized by trituration with 3 mL of MeCN (yield,
1.00 g, th. 1.18 g, 85.0%). An analytical sample was obtained
by two crystallizations from MeCN, mp 156-57 °C (lit.9 mp
153 °C).
Anal. Calcd for C24H25N: C, 88.03; H, 7.69; N, 4.28. Found:
C, 87.97; H, 7.74; N, 4.42.
Detr ityla tion P r oced u r es. For the sake of rate compari-
son, unless otherwise stated, all the detritylations have been
performed at room temperature and stopped, for Brønsted
acids and HgX2 solutions in these acids, with the addition of
water 1 min after complete substrate dissolution. When use
was made of the reductive procedure with HgX2/MeCN, care
has been employed in adding NaBH4 just once the lowest
substrate concentration has been reached in the complexing
solution, that is 5 min after the reaction mixture has become
clear and until TLC monitoring has shown constant chromato-
grams (for trityl-amines only). TLC, with CHCl3 as eluent,
shows disappearance or attenuation of the substrate spot, as
due to complexation along with the presence of a very intense
spot at the origin, due to excess HgX2, the intermediate -onium
complex and the mercury-adduct of the detritylated parent
compound (see eq 1). The relative amount of the -onium
complex and the adduct in the mixture is monitored by the
intensity of the TrX-spot. Presence and intensity of this spot
appears to be related with the yellow color of the substrate
solution in HgX2/MeCN. The presence of the -onium complex
in the spot at 0.0 is indicated by the more or less intense yellow
color developed by this spot in the presence of TFA vapors, as
due to Tr residues. Absence of this chromatic reaction, ac-
companied by lack of the spot of the original substrate, points
to complete detritylation and conversion to the mercury adduct
of the parent compound, as observed in Et-STr. All these
mercury compounds give black spots when sprayed with
NaBH4 in water.
8. N,O-Ditr ityleth a n ola m in e (EA-NOTr ). Ethanolamine
hydrochloride (0.30 g, 3.1 mmol) was added to a mixture of
1.80 g of TrCl (6.5 mmol) and 1.00 g of TEA (9.9 mmol) in 6
mL of MeCN. The suspension was refluxed for 1 h, cooled to
room temperature, and kept at - 4 °C for 2 h. After filtration,
the solid was washed with MeCN, water and once again with
MeCN, dried under vacuum, obtaining 0.90 g of chromato-
graphically pure crystals (th. 1.68 g, 53.57%) with mp 168-
70 °C dec. The analytical sample has been obtained by one
crystallization from MeCN.
Anal. Calcd for C40H35NO: C, 88.04; H, 6.46; N, 2.57.
Found: C, 87.83; H, 6.51; N, 2.61.
9. N-Tr ityleth a n ola m in e (EA-NTr ). This compound has
been obtained by addition of 1.00 g (3.6 mmol) of TrCl to a
large excess of ethanolamine in 5 mL of MeCN. The mixture
was kept under stirring at rt for 0.5 h and then poured in 100
mL of water. The crude solid, obtained by filtration, washed
with water, and dried (0.90 g, th. 1.09 g, 82.6%), was dissolved
in anhydrous Et and precipitated as hydrochloride with
gaseous HCl (platelets, mp 185-87 °C).
Anal. Calcd for C21H22ClNO: C, 74.22; H, 6.52; N, 4.12.
Found: C, 73.95; H, 6.84; N, 4.24.
10. O-Tr ityleth a n ola m in e (EA-OTr ). Excess of ethanol-
amine hydrochloride (6.00 g, 61.5 mmol) was added to a
solution of 6.00 g (21.5 mmol) of TrCl in 20 mL of anhydrous
In a typical experiment, 1 × 10-4 mol of the substrate were
dissolved in 1 mL of the Brønsted acid or 5 mL of the HCl/
MeCN or HgX2/MeCN solution. The metal acid has been
employed in a molar ratio up to 10:1 with respect to the
substrate, to avoid dimerization of the type -D-Hg-D- and to
ensure high equilibrium concentrations of the -onium complex
and of the detritylated mercury adduct. All the operations have
(10) Smith, H. A.; Smith, R. J . J . Am. Chem. Soc. 1948, 70, 2400.

Reductive Demercuration in Deprotection of Trityl Thioethers
J . Org. Chem., Vol. 66, No. 23, 2001 7621
been performed in a nitrogen atmosphere, under stirring and
cooling with ice the reaction vessel.
original tritylated amine. The same results have been obtained
when the hydrochlorides were suspended in HCl/MeCN or
THF. The addition of HgCl2 to the water phase at the end of
the reduction procedure in THF did not produce any precipita-
tion, pointing to absence of Pip.
Detritylation was assumed to be complete when the sub-
strate spot was not visible in the TLC of the final collected
solid or organic solution. Whenever TrOH or TrH have been
isolated, these compounds were determined gravimetrically
and a substantial agreement with chromatographic indications
was obtained. Figures for partial deprotection have not been
determined. Procedures employed in this work are sum-
marized below.
(A) Detr ityla tion w ith P u r e Br øn sted Acid s (TF A or
AcOH, H2O) or Th eir Solu tion s in Ap r otic Solven ts (HCl/
MeCN, H2O). The HCl solution was prepared by bubbling dry
HCl in anhydrous MeCN. Reaction with excess AgNO3 in
water has been employed to determine HCl concentration. The
highest concentration employed was 0.2 M, that is the HgCl+
concentration in the detritylation experiments with the highest
HgCl2/substrate molar ratio. Water was added to the sub-
strate solution in the acidic mixture up to complete precipita-
tion of the solid. The resulting suspension was neutralized or
alkalanized with NaHCO3 or concentrated KOH, then filtered
off or extracted with Et or CHCl3.
(B) Detr ityla tion w ith Solu tion s of HgX2 in Br øn sted
Acid s (HgX2/TF A or AcOH, H2O). In the case of tritylamines
and trityl ethers, once the solution was clear, it has been left
under stirring for 1 min, then water was added, the resulting
suspension was filtered off, and the solid was washed thor-
oughly with water. Trityl thioethers, under these conditions,
gave rise to mercury mercaptides, more or less soluble in
organic solvents, thus interfering with gravimetric determi-
nation of TrOH. In these cases, filtration of the Et solution
through silica gel has led to isolation of pure TrOH.
Free thiols were obtained by cautious addition of sodium
borohydride (portionwise as solid or as concentrated solution
in H2O) to the crude alkaline suspension obtained as described
previously (procedure HgX2/TFA or AcOH, H2O, NaBH4).
Extraction with Et or CHCl3 of the neutral or basic suspension
in which all the borine complexes were destroyed dissolved
TrOH and any organic solid present. The final mixture
consisted of two clear phases, with mercury settled on the
bottom of the reaction vessel. Alternatively, the alkaline
suspension was filtered off, and the solid was washed with
water, dried, and reduced by NaBH4/MeCN.
(C) Red u ctive Detr ityla tion w ith Solu tion s of HgX2 in
MeCN (HgX2/MeCN, Na BH4). The substrate was dissolved
in HgCl2/MeCN, and the solution was left under stirring up
to constant chromatogram. Then, solid NaBH4 was added
cautiously portionwise to the solution, obtaining precipitation
of a gray, dense solid releasing bubbles of gas. Gradually, the
solid darkened and settled (this same behavior characterizes
the reaction of NaBH4 with pure HgCl2 and Hg(OAc)2). At this
stage, whenever complete detritylation occurred, stoichiometric
TrH was recovered from the solution (see, for example, Et-
OTr in detritylation experiments section) and both the sub-
strate and its parent compound have disappeared. The end of
reaction has been assumed when gaseous bubbling was ceased.
Small portions of water were added cautiously to the
suspension of the gray solid, giving rise to intense production
of gas. Then, 1:1 HCl was dropped into the reaction mixture
up to complete destruction of borine complexes and Et was
added, for the sake of obtaining two clear phases, with mercury
settled on the bottom of the reaction vessel. Further addition
of concentrated KOH or AcONa led to regeneration of organic
bases and their extraction in the organic phase.
Examples of reductive demercuration of adducts of thiols
and of complexes of amines are as follows.
Red u ctive Dem er cu r a tion of Ch lor om er cu r y(II) Hy-
d r oxyeth ylm er ca p tid e w ith Na BH4/MeCN. The substrate
ME-SHgCl has been prepared by adding under stirring 1.0 mL
(0.0128 mol) of ME in 20 mL of EtOH to a hot solution of 4.0
g (0.0148 mol) of HgCl2 in 150 mL of EtOH. When the addition
of thiol was completed, the obtained suspension was refluxed
for 5 min, then left aside for 1 h and filtered off. The solid was
washed with EtOH and Et (3.66 g, th. 4.00 g, 91.5%). The
analytical sample was obtained by crystallization from EtOH,
in the presence of a small amount of HgCl2, mp 161-163 °C
(lit.11 mp 135-140 °C).
Anal. Calcd for C2H5ClOSHg: C, 7.67; H, 1.61; S, 10.24.
Found: C, 7.45; H, 1.63; S, 9.88.
In three experiments, solid NaBH4 was added portionwise
under stirring to a suspension of about 200 mg of ME-SHgCl
in 5 mL of MeCN. Weighting of the produced Hg0 has
furnished an average value for Hg % of 63.82 (th. 64.05).
Red u ctive Dem er cu r a tion of Tw o HgCl2 Com p lexes of
P ip . (a) The addition of 0.182 mL of Pip (1.8 mmol) to a
solution of 1.00 g of HgCl2 (3.7 mmol) in 5 mL of MeCN has
given rise to precipitation of a microcrystalline solid with mp
161-63 °C and composition corresponding to Pip‚HgCl2. Anal.
Calcd for C5H11Cl2HgN: C, 16.84; H, 3.11; N, 3.93. Found: C,
16.74; H, 3.35; N, 3.87.
(b) A well-crystallized compound of formula Pip‚HgCl2‚HCl
was prepared by addition of 2.33 g of HgCl2 (8.6 mmol) to a
suspension of 1.00 g of piperidine hydrochloride (8.2 mmol) in
5 mL of a 0.7 M solution of HCl in MeCN. The resulting clear
solution was kept at 0 °C for 4 h and then filtered off. The
crystalline solid, washed with the cold HCl/MeCN solution and
ether, was dried under vacuum and employed as analytical
sample (0.93 g, th. 3.37 g, 27.6%), mp 108-9 °C. Anal. Calcd
for C5H12Cl3HgN: C, 15.29; H, 3.08; N, 3.58. Found: C, 15.42;
H, 2.97; N, 3.48.
Samples of both these compounds were reduced with NaBH4
as suspension in THF or solution in H2O, obtaining Hg0. Free
Pip has been revealed in water solution with ninhydrin or with
HgCl2, obtaining a white precipitate at pH 7.
Detr ityla tion Exp er im en ts. The most representative
detritylation experiments are reported below (see also, Figures
1-3).
Ca ta lyzed a n d Un ca ta lyzed Red u ctive Detr ityla tion
of Tr Cl, Tr OAc, a n d Tr OH to Tr H. TrCl and TrOAc12 were
quantitatively and rapidly reduced to TrH by portionwise
addition of solid NaBH4 to their MeCN solutions.13 TrOH
remained unaffected by this same treatment, but more or less
relevant quantities of TrH were produced by addition of NaBH4
to solutions of this substrate in HCl/MeCN or TFA and vice
versa.
Solutions of TrCl in HgCl2/MeCN are reduced by NaBH4 to
TrH. TrOH is reduced to TrH if NaBH4 is added portionwise
to an its solution in HgCl2/MeCN and vice versa. The use of
Hg(OAc)2 in these same procedures has left this substrate
unaltered.
Detr ityla tion of Mon ofu n ction a l Su bstr a tes. Detr ityl-
a tion of P ip -Tr w ith HgCl2/MeCN, Na BH4. On adding Pip-
Tr to HgCl2/MeCN, the colorless mixture became rapidly
turbid. TLC taken 5 min after mixing showed an intense
substrate spot. Reduction of this solution with NaBH4 gave
Rea ctivity of Na BH4 w ith Isola ted Sp ecies in Equ ilib-
r iu m (1). With the only exception of TrCl and TrOAc (see
below), all the tritylated substrates investigated in this work
have been lost unaltered by treatment with NaBH4 in any
protic or aprotic solvent.
(11) Bennett, G. K. J . Chem. Soc. 1922, 2139.
(12) TrOAc has been prepared in situ by mixing equimolar AgOAc
and TrCl in anhydrous MeCN.
(13) The NaBH4 reduction of halides capable of forming stable
carbenium ions is a well-known reaction but it is reported to be very
slow in absence of water. See: (a) Brown, H. C.; Bell, H. M. J . Org.
Chem. 1962, 27, 1928. (b) Bell, H. M.; Brown, H. C. J . Am. Chem. Soc.
1966, 88, 1473.
To clarify the catalytic effect of protons on the nucleophilic
attack of the hydride ion on a tritylamine, the hydrochlorides
of Pip-Tr and EA-NTr have been treated with NaBH4. The
addition of the reactant to the MeCN or THF solution of these
hydrochlorides gave rise to production of hydrogen and dilution
with water of the mixture produced only recovering of the

7622 J . Org. Chem., Vol. 66, No. 23, 2001
Maltese
back Pip-Tr. When the mixture was left aside for 48 h, a
considerable amount of a white solid was produced and Pip-
Tr disappeared from TLC of the solution. The solid, filtered
and washed with cold MeCN and Et, had a mp 131-133 °C
dec. The elemental analysis pointed to Pip-Tr‚6HgCl2 as its
simplest molecular formula (Anal. Calcd for C24H25Cl12Hg6N:
C, 14.73; H, 1.29; N, 0.27. Found: C, 15.03; H, 1.30; N, 0.39).
Addition of NaBH4 to the suspension or either to the isolated
solid or the filtrate led to mixtures of Pip-Tr and TrH. The
presence of Pip in water has been revealed by addition of
HgCl2, which gave rise to precipitation of a white insoluble
solid.
Detr ityla tion of EtO-Tr w ith HgCl2/MeCN, Na BH4. (a)
A colorless solution of 0.0997 g (0.35 mmol) of Et-OTr in 15
mL of a MeCN solution of 0.9453 g (3.5 mmol) of HgCl2 was
reduced with NaBH4 just 5 min after mixing. TLC monitoring
before reduction showed permanence of the substrate spot and
lack of TrCl. The 0.0 intense spot did not develope any color
when exposed to TFA vapors. After the rapid addition of
NaBH4, the substrate disappeared from TLC and only TrH
was detected in the eterogeneous solution. Water was added
2 min after the addition of the reductant was completed. At
the end of the treatment, the TLC of the organic layer, with
system B as eluent, showed only the TrH spot. Evaporation
of the organic layer has given 0.0789 g (th. 0.0845 g) of crystals
with mp 91-94 °C (TrH). (b) In a second experiment on 0.1050
g of Et-OTr, the suspension obtained after the addition of
NaBH4, after the dark-gray solid was settled, was filtered off
and the solid washed with MeCN and Et. Water was added to
filtrates, and after destruction of the borine complexes, the
evaporation of the organic layer gave 0.0810 g of TrH, 91.1%.
(c) In a third experiment, the Et-OTr solution in HgCl2/MeCN
was left aside for 4 days, then treated as before, obtaining
similar results.
Detr ityla tion of Et-STr w ith HgCl2/MeCN, Na BH4. (a)
A yellow solution of 1 × 10-4 mol of Et-STr and 1 × 10-3 mol
of HgCl2 in 3 mL of MeCN has given TLC(CHCl3) showing a
very weak substrate spot and an intense one due to TrCl; the
0.0 spot remains almost colorless on exposure to TFA vapors.
After addition of NaBH4 and settling of the dark-gray solid,
the suspension was treated as described above for Et-OTr
under (a) and (b), obtaining in both cases quantitative recover-
ing of TrH. (b) A solution of 1 × 10-4 mol of both Et-STr and
HgCl2 in 3 mL of MeCN has given a TLC(CHCl3) with spots
at 0.0, 0.6 (TrCl) and 0.9 (Et-STr). (c) In a further experiment,
1 × 10-4 mol of this same substrate were dissolved in a solution
of 1 × 10-3 mol of Hg(OAc)2 in 3 mL of MeCN. The most
relevant difference with respect to the results obtained with
HgCl2 was the presence in TLC(CHCl3) of a tail 0.0-0.6, which
became yellow with TFA vapors. The other results were similar
to those obtained with HgCl2.
Detr ityla tion of Bifu n ction a l Mon otr ityla ted Su b-
st r a t es. (i) S-Det r it yla t ion of ME A-STr t o ME A w it h
Hg(OAc)2/AcOH, H2O (See Figure 1). A solution of 32.0 mg
of MEA-STr (1 × 10-4 mol) and of an excess of Hg(OAc)2 in 1
mL of AcOH was stirred for 5 min at room temperature, and
then water was added (10 mL). The suspension was filtered
off, and the solid was washed with water and dried (TrOH
chromatographically pure, 22.3 mg, th. 26.0, 85.8%). The
filtrate, treated with NaBH4, HCl, and NaHCO3 up to pH 7-8,
was titrated with 0.45 × 10-4 mol of I2. Care was taken in
controlling the complete destruction of NaBH4. In a blank, an
equal quantity of the reductant was dissolved in water and
the solution treated as described above: the final mixture did
not reduce I2/H2O or a HgCl2 solution in water.
of the solid residue with TFA, H2O led to disappearance of
the spots due to MEA-NTr and MEA-NSTr and to permanence
of that due to MEA-STr. Formation of free MEA has been
excluded, since the water phase did not react with I2, nor give
any chromatic reaction with Ni(OAc)2. Treatment of the same
residue with HgX2/TFA, H2O and addition of NaBH4 to the
water suspension, leading to complete detritylation of all the
MEA-derivatives, gave rise to a water solution with an 80%
of the expected reducing power.
(iii) S-Detr ityla tion of ME-STr to ME w ith Hg(OAc)2/
AcOH, H2O (One-Pot Procedure, See Figure 2). A solution of
32.0 mg of ME-STr (1 × 10-4 mol) and of an excess of mercury
acetate in 1 mL of AcOH was stirred at room temperature for
5 min, then 10 mL of water was added and the resulting
suspension was brought to pH 7-8 and reduced with NaBH4.
At the end of the treatment, Et was added and the suspension
was extracted three times. Drying and evaporation of the Et
solution gave 21.0 mg of TrOH (th. 26.0 mg, 80.8%) and the
water solution was rapidly titrated with 4.1 mL of I2/H2O 0.01
M (0.41 × 10-4 mol), according to Zervas and Photaki.2,3
(iv) La ck of Detr ityla tion of ME-STr w ith TF A, H 2O.
A yellow solution of ME-STr in TFA, kept under stirring at
room temperature for 5 min, diluted with water 1:10 gave a
sticky solid whose TLC points to the prevalent presence of the
original compound, with TrOH and S-trityltrifluoroacetyl-
mercaptoethanol as impurities.
(v) O-Detr ityla tion of ME-OTr to ME w ith HgX2/TF A,
H2O, Na BH4 (See Figure 2). As in the case of MEA-NTr, ME-
OTr was not isolated as such and the effects of detritylation
procedures have been merely monitored by TLC. With CHCl3
as eluent, ME-OTr has Rf ) 0.85 and its spot becomes brown
when sprayed with Ni(OAc)2/MeOH. The crude solid obtained
from S-detritylation of ME-OSTr cannot be O-detritylated with
TFA, H2O, because this treatment gives rise to migration. Free
ME from ME-OTr has been obtained only by the procedure
HgX2/TFA, H2O, NaBH4.
(vi) Detr ityla tion of EA-NTr w ith HgCl2/MeCN, Na BH4.
The TLC(CHCl3) of a colorless solution of EA-NTr in HgCl2/
MeCN showed three spots overlapped at the origin. This
solution was reduced five minutes after its formation and the
result was recovering of unaltered EA-NTr. Another similar
solution, left aside for 48 h, besides the two intense spots due
to HgCl2 and the complex, showed spots due to TrCl and EA-
NOTr, this latter recognized by a chromatic reaction with
ninhydrin. Reduction with NaBH4 produced EA-NTr and TrH.
Lack of EA-OTr in the mixture before reduction was revealed
in the following way. One milliliter of this solution was poured
in a solution of KOH/H2O 2 M and the resulting suspension
was extracted with Et. TLC(CHCl3) of the organic layer,
besides TrOH and EA-NOTr, revealed only EA-NTr. In
contrast, EA-NTr was rapidly detritylated by the TFA, H2O
procedure, with quantitative recovery of TrOH.
(vii) O-Detr ityla tion of EA-OTr to EA w ith HgCl2/
MeCN, Na BH4 (See Figure 3). EA-OTr was easily deprotected
by the procedure HgCl2/MeCN or THF, NaBH4, with traces of
migration and formation of EA-NTr and EA-NOTr. TrH was
the secondary product.
Detr ityla tion of Bifu n ction a l Ditr ityla ted Su b-
str a tes: MEA-NSTr (Figure 1). (i) S-Detr ityla tion of MEA-
NSTr to MEA-NTr w ith HgCl2/MeCN, H2O, Na BH4. A 0.20
g (0.4 mmol) portion of the substrate was dissolved in a
solution of 0.90 g of HgCl2 (3.3 mmol) in 8 mL of MeCN. The
yellow solution readily gave rise to precipitation of a white
solid. TLC monitoring indicated presence of TrCl and absence
of substrate. Water was added to this suspension up to solid
dissolution and reprecipitation. Portionwise addition of solid
NaBH4 and treatment as reported above, gave rise to two clear
phases. The organic phase, equilibrated against the water
solution at pH ) 8, was isolated, dried on Na2SO4, and
evaporated under vacuum. Et was added to the solid residue
and the Et solution was treated with iodine according to the
previously described procedure (see substrates, no. 7), giving
rise to 0.08 g of chromatographically pure CYA-2Tr (th. 0.11
g, 72.7%).
(ii) N-Detr ityla tion of MEA-NTr to MEA-STr a n d to
MEA (See Figure 1). Owing to its instability, this substrate,
obtained by detritylation of MEA-NSTr with the procedure
described below, has not been isolated and its behavior under
detritylation has been monitored only by TLC. Its Rf(CHCl3)
is 0.75 and the spot becomes red-brownish when sprayed with
Ni(OAc)2/MeOH. In the organic phase obtained at the end of
the MEA-NSTr detritylation, MEA-NTr is always accompanied
by MEA-NSTr, MEA-STr and TrOH as impurities. Once the
organic solution has been distilled under vacuum, treatment

Reductive Demercuration in Deprotection of Trityl Thioethers
J . Org. Chem., Vol. 66, No. 23, 2001 7623
(ii) S-Det r it yla t ion of ME A-NSTr t o ME A-NTr w it h
Hg(OAc)2/MeCN, Na BH4. Starting from the same quantity
of the substrate, CYA-2Tr have been obtained, with a higher
yield (81.8%) with respect to the treatment with HgCl2. This
result can be ascribed to the fact that in the HgCl2 procedure,
the strong acid produced during the NaBH4 addition catalyzes
Tr-migration from N- to S-atom.
(iii) NS-Detr ityla tion of MEA-NSTr th r ou gh MEA-
SHgX w ith HgX2/TF A, H2O, Na BH4 (One-Pot Procedure).
When water was added to a yellow solution of MEA-NSTr in
HgX2/TFA, a solid was obtained (a mixture of MEA-SHgX and
TrOH, by TLC) which has been reduced with NaBH4 directly
in the suspension, once the solution was neutralized. After
reduction, 91% of free thiol was detected with I2 titration of
the water phase. An analogous result has been obtained on
isolating the solid mixture and reducing it with NaBH4/MeCN.
Detr ityla tion of Bifu n ction a l Ditr ityla ted Su b-
str a tes: ME-OSTr (Figure 2). (i) S-Detr ityla tion of ME-
OSTr to ME-OTr w ith Hg(OAc)2/MeCN, Na BH4 (One-Pot
Procedure). A 0.5610 g portion of ME-OSTr (0.10 mmol) was
added to a suspension of excess Hg(OAc)2 in 6 mL of MeCN,
and the mixture was left under stirring for 5 min, then solid
NaBH4 was added portionwise. At the end of the treatment,
the organic phase was rapidly titrated with I2/Et in the
presence of a small amount of a concentrated solution of
NaHCO3: 0.04 mmol of I2 was employed, yield 88.0%. After
separation and washing with water, the organic solution was
dried over Na2SO4 and then evaporated to dryness (0.0280 g,
th. 0.0320, 87.5%). The disulfide has been crystallized from
MeCN.
(ii) O-Detr ityla tion of ME-OSTr to ME-STr w ith TF A
a n d HCl/MeCN, H2O. Quenching by water of a yellow
solution of ME-OSTr in TFA led to mixtures of unreacted ME-
OSTr, ME-STr, and TrOH. Addition of water to an uncolored
solution of this same substrate in a moderately concentrated
solution of HCl in MeCN led to rapid detritylation to ME-STr.
(iii) OS-Detr ityla tion of ME-OSTr to ME-SHgX w ith
Hg(OAc)2/ TF A, H2O. The procedure was as above for N,S-
detritylation of MEA-NSTr to MEA-SHgX with HgX2/TFA,
H2O. The yield in thiol was 86% in the one-pot procedure (see
below) and less when the mercaptide was isolated and reduced
in MeCN suspension.
(iv) OS-Detr ityla tion of ME-OSTr to ME w ith HgX2/
TF A, H2O, Na BH4 (One-Pot Procedure). ME-OSTr was dis-
solved in TFA containing excess Hg(OAc)2, giving a yellow
solution. After addition of water and neutralization, the
suspension was reduced with NaBH4, giving rise to two clear
final phases. The organic phase gave almost quantitative
recovery of 2 equiv of TrOH. The Ni(OAc)2/MeOH test was
negative for the organic layer. In contrast, addition of Ni(OAc)2/
H2O to the inorganic layer at pH 8 produced an intense red-
brownish color.
Detr ityla tion of Bifu n ction a l Ditr ityla ted Su b-
str a tes: EA-NOTr (Figure 3). (i) O-Detr ityla tion of EA-
NOTr w ith HgCl2/MeCN, Na BH4. The colorless solution of
0.1046 g of EA-NOTr in excess HgCl2/MeCN, 5 min after its
formation, had TLC(CHCl3) showing absence of TrCl and
permanence of the uncomplexed substrate (however, spots at
0.0 became yellow in the presence of TFA vapors). Once this
solution was reduced with NaBH4, the final alkaline water
suspension was thoroughly extracted with Et, and after drying,
the extracts were distilled under vacuum. The solid residue
was dissolved in Et and the resulting solution was filtered off.
Bubbling of dry HCl produced 0.0554 g of EA-NTr‚HCl (th.
0.0652 g, 84.9%) which was isolated by filtration. EA-NTr was
obtained by extraction with concentrated KOH of an Et
suspension of this solid, and recognized by TLC. The filtrate,
after solvent evaporation, has given off 0.0415 g of TrH (th.
0.0468 g, 88.7% for a detritylation of only one group).
Discu ssion
In principle, the addition of a nucleophile stronger than
X- to the equilibrium mixture (1) favors the formation
of the deprotected substrate R-D-A by attack on tri-
phenylcarbenium ion, on TrX and on the Tr-residue of
the intermediate complex (R-D+(A)-Tr)X-, this last spe-
cies appearing to be more suitable for substitution than
R-D-Tr itself for having a better leaving group. When :Nu
) H2O or NaBH4, TrOH and TrH are expected as final
secondary products, respectively. A further factor favor-
ing deprotection, when A+ is HgX+ and NaBH4 is
employed as nucleophile, is the reducibility of R-D-HgX
to R-D-H. Competition of the nucleophile and the inter-
mediate complex for the Lewis acid A+, has to be taken
into account as unfavorable factor in detritylation, be-
cause abstraction of A+ from the complex produces not
reducible R-D-Tr.
As previously said, the yellow color assumed in many
instances by the catalyzed mixtures indicate an high
degree of shifting of equilibrium (1) toward R-D-A and
TrX, for it being due to free Tr+ or ionic couples with
triphenylcarbenium ion.7,14 Whenever this color is seen,
its formation rate is always high, pointing to rapid
reactions producing and consuming species taking part
to eq 1. Further information can be obtained on observing
the quenching effects produced on these colored solutions
by addition of the nucleophile. The first one is that
quenching not always accompanies detritylation. Mono-
functional trityl ethers, -amines and -thioethers give rise
to deeply yellow TFA solutions, but rapid addition of
water to these produces only O- and N-detritylation. It
can be guessed that only in the first two cases loss of
color is due to the transformation of Tr+ and TrX into
TrOH, while for a trityl-thioether it is apparently an
effect of reversion of equilibrium (1) toward R-D-Tr. The
second information is that the irreversible production of
TrOH from Tr+ and TrX cannot be the only driving force
of detritylation, otherwise, trityl-thioethers as well should
be detritylated by TFA, H2O. The third information is
that, since fading is rapid, reactions with water leading
to quenching are fast.
Of these same monofunctional substrates, only Et-STr
gives rapidly rise to a yellow solution when dissolved in
HgCl2/MeCN. Addition of water to the solution of Et-STr
gives rise to rapid fading and to S-detritylation, while
this same treatment of the Pip-Tr and of the Et-OTr
solutions leads essentially back to the unreacted sub-
strates. In this case, as well, detection of Tr+ does not
ensure detritylation. Furthermore, in S-detritylation,
rates of reactions in equilibrium (1) and of those between
species in eq 1 and water are rapid. Furthermore,
solutions of the same substrates in Hg(OAc)2/MeCN or
AcOH are colorless, but, addition of water leads to
detritylation of trityl-thioethers alone. Hence, the ef-
ficiency of the procedure does not depend on the degree
of shifting of eq 1.
In the procedure TFA, H2O, proton exchange between
H2O and A+X- seems to be ineffective on detritylation;
otherwise, trityl ethers, trityl amines, and trityl thio-
ethers should exhibit the same behavior. Analogous
consideration can be extended to cases in which A+
)
HgX+. Thus, neither the nucleophilic attack onto Tr+ or
(ii) O-Detr ityla tion of EA-NOTr w ith HCl/MeCN, H2O.
EA-NOTr, sparingly soluble in MeCN, dissolved rapidly in the
presence of HCl, soon giving a suspension of the hydrochloride
of EA-NTr from which EA-NTr has been recovered by treat-
ment with Et/concentrated KOH.
(14) (a) Evans, A. G.; J ones, J . A. G.; Osborne, G. O. Trans. Faraday
Soc. 1952, 50, 16. (b) Bayles, J . W.; Evans, A. G.; J ones, J . R. J . Chem.
Soc. 1955, 206.

7624 J . Org. Chem., Vol. 66, No. 23, 2001
Maltese
TrX nor competition of H2O and A+X- for A+ determine
the result of detritylation procedures such as TFA, H2O
and HgX2/MeCN, H2O.
as nucleophile. Trityl ethers and trityl amines are de-
tritylated by HgX2/TFA, H2O, but, since they remain
unaffected by treatment with HgCl2/MeCN, H2O, these
results indicate that detritylation is due to the strength
of the Brønsted acid employed.
In the H+-catalyzed procedures employing H2O as
nucleophile, of the remaining two factors affecting de-
tritylation, competition of H2O and (R-D+(H)-Tr)X- for
H+ seems to be the most effective in explaining the
reported behavior of trityl ethers, trityl amines, and trityl
thioethers. In general, the attack of the nucleophile on
the intermediate onium-complex can produce both A+-
and Tr+-abstraction, according to the following equilib-
rium:
On passing to the use of NaBH4 as nucleophile, the
comparison of the results obtained on adding the reduc-
tant to HgCl2/MeCN solutions of Et-STr and Et-OTr may
elucidate many questions. The former substrate gives a
yellow solution, the latter a colorless one. TLC-monitoring
confirms that Et-STr is practically absent in the equi-
librium mixture, while Et-OTr is negligibly complexed.
However, both are completely deprotected by NaBH4,
thus pointing to the fact that the reactions producing TrH
and consuming Tr+, TrCl and R-D-HgCl goes to comple-
tion just because reactions in equilibrium (1) are fast
when D ) S or O. Partial detritylation of Pip-Tr allows
to experience a different situation. When Pip-Tr is in
equilibrium in HgCl2/MeCN, lack of free substrate and
presence of only traces of TrCl in the heterogeneous
mixture should indicate that the prevalent species is the
complex. Furthermore it cannot be said if the production
of Tr+, TrCl and PipHgCl is fast, but it can be supposed
as such, on grounds of steric repulsion among the four
bulky groups in the complex. Thus, the production of
mixtures of Pip-Tr and TrH by addition of NaBH4 has to
be ascribed to competition between the production through
eq 1 of species reducible to TrH and the reduction of the
ammonium-complex. Furthermore, it should be remem-
bered that the complex can be reduced through two
different routes. The first one, in analogy with the
reduction of PipTr ‚ HCl, leads to Hg0 and Pip-Tr, the
second one produces first TrH and Pip-HgCl, which, in
turn, is reduced to Pip and Hg0. This latter is a second
route to detritylation, but its contribution to the produc-
tion of TrH appears quite dubious.
R-D-Tr + A-Nu+ h R-D+(A)-Tr + :Nu h
R-D-A + Tr-Nu+
If, as reasonably expected, in eq 2 proton abstraction
is much faster than detritylation, the deprotonation
equilibrium is rapidly reached. If the affinity of the
nucleophile for A+ is higher than that of the complex, an
overall reversion of equilibrium (2) toward R-D-Tr should
be expected. In contrast, if the D+-A bond is preserved,
the trityl-residue can be removed by attack on Tr+ and
TrX. In other words, when A+ ) H+, the equilibrium
concentrations are determined by the relative affinity of
D and :Nu for H+, this decreasing along the series N >
O > S. Consistently, in the H+-catalyzed detritylation
procedures, nucleophiles having oxygen as donor atom
(water or alcohols) cannot compete with nitrogen for
protons, so that the addition of water to TFA solutions
of trityl amines gives rise to trityl-displacement. In
contrast, the more acidic S+-H moieties easily lose
protons, so that water addition to TFA solutions of trityl
thioethers causes recovering of the unaltered substrate.
Addition of mercaptans to TFA solutions of trityl thio-
ethers gives rise to equilibria based on the trityl migra-
tion from the original to the added mercaptan. Accord-
ingly, the presence of a free thiol group in the molecule
of a trityl ether or trityl amine gives rise to H+-catalyzed
Tr migration, as observed for ME-OTr and MEA-NTr. On
passing to ditritylated bifunctional substrates, proce-
dures TFA, H2O and HCl/MeCN, H2O give results in
substantial agreement with those reported above for
monotritylated substrates.
Different possible mechanisms can be proposed for the
reduction of TrX (X ) -Cl or -OAc). In the SN1
mechanism, the first step consists of the production of
-
triphenylcarbenium ions and BH4 acts merely as a
scavenger of these ions.17 A second mechanism requires
an SN2 backside attack on TrX, while a third one, a four-
center mechanism, requires the presence of a polar bond
in the molecule and proceeds through attack of the
-
nucleophilic hydrogen of BH4 onto the electrophilic
Finally, as the last factor favoring detritylation, nu-
cleophilic substitution of H2O on Tr-groups of the complex
itself, for steric reasons, should have a rate lower than
attacks on Tr+ and TrX. However, contribution of this
reaction to detritylation cannot be excluded, whenever
proton exchange between :Nu and (R-D+(H)-Tr)X- allows
significant equilibrium concentrations of the complex.
Metal acids, introduced in a solution of a trityl-
derivative in a protic solvent, compete with protons for
the central atom in the -onium complexes and overwhelm
protons themselves as acceptors when the solvent is a
weak Brønsted acid. The nucleophile added to these
solutions and, of course, to those in aprotic solvents, has
to compete with the -onium complex for A+ ) HgX+, a
soft metal acid15 for which the affinity scale to be followed
should be S > N > O.16 So, the introduction of HgX2 into
a TFA, AcOH, or MeCN solution of any S-trityl deriva-
tive, leads invariably to the isolation of the metal
mercaptide and triphenylcarbinol, if water is employed
center, while the borine atom accepts electronic charge
coming from the nucleophilic center in the bond.18 If this
last would be the mechanism, lack of detritylation by
simple attack of NaBH4 on the other tritylated substrates
under examination in this work should be referred to the
low polarity of the D-Tr bond. However, besides bond
polarity, steric hindrance due to the bulky Tr-moiety
should prevent easy formation of four-center activated
states. Even in case of TrCl, so structurally simple with
respect to the other trityl-derivatives investigated, it
appears more conceivable a backside attack on Tr- or,
better, quenching of Tr+, rapidly produced by the sub-
strate itself. Furthermore, although charge separation
in TrOH (and in the remaining tritylated substrates)
should not be very different with respect to that in TrCl
or TrOAc, NaBH4 in pure MeCN leaves unaltered TrOH.
In the reaction of TrX with NaBH4, instead of bond
(17) This mechanism is assumed as predominant by Olah (ref 6) in
the hydrogen transfer reaction between NaBH4 and Tr+PF6-, MeCN
being a solvent favouring the formation of Tr+.
(15) Pearson, R. G. J . Chem. Educ. 1968, 45, 581 and 643.
(16) Sigel, H.; McCormick, D. B. Accounts Chem. Res. 1970, 3, 201.
(18) Dessy, R. E.; Paulik, F. J . Chem. Educ. 1963, 40, 185.

Reductive Demercuration in Deprotection of Trityl Thioethers
J . Org. Chem., Vol. 66, No. 23, 2001 7625
polarization, it seems the nature of good leaving groups
of Cl- and OAc- to determine production of TrH. Hence,
the four-center mechanism does not appear a conceivable
route for cleaving bonds with trityl groups.
ME-OTr, leads to recovery of ME-OTr. This result seems
to convincingly support the four-center model for the
reduction mechanism. Both these substrates in HgCl2/
MeCN are essentially solutions of ME-OTrSHgCl, a
mercaptide in which the O-atom, for its â-position with
respect to S, should be strongly coordinated to the metal
atom, that is its lone pair is strongly involved in this bond
with the only metal atom present in the molecule. When
NaBH4 is employed, electrophilic borine prefers the much
more nucleophilic center on sulfur with respect to that
on oxygen, thus leaving uncleaved the O-Tr bond.
Both complexation with H+ or HgCl+ increases charge
separation in the Tr-D bond, transforming R-D+ in a
leaving group better than in the uncomplexed substrate.
Thus, nucleophilic substitution by hydride-ions could be
expected, as for TrX. In contrast, in these cases the
experimental evidence indicates that the Tr-N+ bond is
not cleaved. The reaction with NaBH4 of the hydro-
chlorides of Pip-Tr and EA-NTr produces H2 instead of
TrH. For hydrochlorides, this result supports the con-
The four-center model for the transition state in the
NaBH4 reduction receives further support by reduction
of HgX2 and R-D-HgX. In the first case, the gray to black
solid produced by addition of NaBH4 to a solution of HgX2
in MeCN should be the four-center complex itself, which
slowly decomposes producing the unstable hydride H-Hg-
X, which, in turn, gives rise to formation of HX and Hg°,
according to well-established studies.22,23 In reduction of
the adducts R-D-HgX the substrate has two different
nucleophilic sites to be approached by the electrophilic
counterpart -BH3-, the X atom or the donor atom D.
Approaching to X should prevent any HX to be released
and induces formation of R-D- ions, while approaching
to D gives rise to production of free HX and leads to the
formation of borine-complexes of the type ((RD)xHy)B-,
which in a further step, in the presence of water,
decompose giving free thiols, amines or alcohols, as
effectively observed.
certed-type mechanism already proposed for the reaction
19,20
of NaBH4 with weak acids
and, in particular, with
trimethylammonium ion. 21 In the tritylammonium com-
plexes investigated in this work, the electrophilic center
approached by the hydride ion is the hydrogen-atom and
never the central carbon atom of the Tr-group. Thus, the
concerted-type mechanism does not allow any detrityl-
ation to occur by attack of the reductant on the inter-
mediate ammonium-complexes. As shown above, this
statement can be extended to intermediate HgX+-
ammonium complexes of tritylated amines, with the only
possible exception of Pip-Tr.
Et-OTr and Et-STr, as examples of substrates where
the presence of a lone pair on D+ allows bonding with
electrophilic borine and the concerted-type mechanism
to appear more suitable than for ammonium-complexes,
are fully detritylated by HgCl2/MeCN, NaBH4. In con-
trast, the same procedure, when applied to ME-OSTr or
J O0156971
(19) Davis, R. E.; Bromels, E.; Kibby, C. L. J . Am. Chem. Soc. 1962,
84, 885.
(20) Dessy, R. E.; Grannen, E. J . Am. Chem. Soc. 1961, 83, 3953.
(21) In this case, however, the four-center complex does not appear
to be consistent with the experimental results; see: Davis, R. E. J . Am.
Chem. Soc. 1961, 84, 892.
(22) Bordwell, F. G.; Douglass, M. L. J . Am. Chem. Soc. 1966, 88,
993.
(23) Barbaras, G. D.; Dillard, C.; Finholt, A. E.; Wartik, T.; Wilzbach,
K. E.; Schlesinger, H. I. J . Am. Chem. Soc. 1951, 73, 4585.