-
Synthesis, DelocPlization and Reactivity in Stable
Diaminocarbenes
Shilpi Gupta
A thesis subrnitted in conformity with the requirements for the
degree of Master's of Science Graduate Department of Chemistry
University of Toronto
@ Copyright by Shilpi Gupta (1 999)
-
National Library l*l of Canada Bibliothèque nationale du Canada
Aquisitions and Acquisitions et Bibliographie Services seMces
bibliographiques 395 Wellington Street 395, rue Wellington OttawaON
K1AON4 Ottawa ON KI A ON4 Canada Canada
The author has granted a non- exclusive licence allowing the
National Library of Canada to reproduce, loan, distribute or sel1
copies of this thesis in rnicroform, paper or electronic
formats.
L'auteur a accordé une Licence non exclusive permettant à la
Bibliothèque nationale du Canada de reproduire, prêter, distribuer
ou vendre des copies de cette thèse sous la forme de
microfiche/film, de reproduction sur papier ou sur format
électronique.
The author retains ownership of the L'auteur conserve la
propriété du copyright in this thesis. Neither the droit d'auteur
qui protège cette thèse. thesis nor substantial extracts fiom it Ni
la thèse ni des extraits substantiels may be printed or othewise de
celle-ci ne doivent être imprimés reproduced without the author's
ou autrement reproduits sans son permission. autorisation.
-
Synthesis, Delocalbtion and Reactivity of Stable Dmrninocarbenes
Master's of Science, 1999 Shilpi Gupta Department of Chemistry,
University of Toronto
Abstract
The synthesis and reactivity of stable diaminocarbenes has been
investigated. The
synthetic routes utilized were reductive dehydrosulhuization of
tetra-substituted thioureas and 1,l-
elimination of HCI frorn carbenium salt, [N2CH]+ CI-. Synthesis
of sterically crowded thioureas
suffers from low yields. The dehydrosulfurization of aminals has
been discovered as a new one-
step synthesis for thioureas and carbenium salts.
Dehydrosulfurization was also investigated for urotropin and
1,3,5-trialkyl-hexahydro-
sym-triazines (investigated as precursors for polycarbenes). The
dehydrosulfurization of sym-
triazines gave ring degradation products and [C2H2(NR)2CH]+
SCN-. Reaction of carbenes and
analogs with alcohols and alkoxides was investigated. The
aromatic 61r-delocalization in carbenes
and related heterocycles was studied at the RHF 1 6-3lG* and
B3LYP / 6-3 IG* level. The
obtained Lowdin bond orders correlate with the aromatic ring
cument (IH-NMR) making them an
excellent computational tool to study the extent of ammatic
delocalization.
-
- Table of Contents - A bstrac t
List of Tables
List of Figures
Ab breviations
Acknowledgments
Chapter 1
1.1
1.1.0
1.2
1.3
1.4
Chapter 2
2.1
2.1.1
2.1.2
2.1.3
2.2
2.3
2.4
2.4.1
2.5
2.5.1
2.5.2
2.5.3
Introduction
Carbenes and Carbenium Ions
S ynthesis
Oxidative Addition of Alkoxides and Alcohols
Metal-Carbene Complexes
Thioureas and Thiourea Derivatives
Results and Discussion
Dehydrosulfurization of Arninals (RzN)2CH2 with Sa
Applications of Thioureas
Objectives
Mechanism and Product Distribution
l=S via [l-R] CI
Synthesis of 3=!3
Synthesis of Imidazolium Salts
Deprotonation of [l-H] CI to give 1
Properties of Imidazolium Salts
Basicity of Carbenes and Acidity of Imidazolium Salts
Ion pairing
Hy bridization
ii
vii
viii
xi
xii
-
Solubilities of Carbenium Salts
Deprotonation Strategy to give 2 from [2-H] SCN
Deprotonation Strategy to give 3 from [3-LI] SCN
Other Deprotonation Bases
Aromatic, Anti-Aromatic and Linear Conjugated
Pol ycarbenes
1,3,S-tn-te~-butyl-hexahydro-sync-triazine with S8
Multistep approaches for the synthesis of Il-'Bu
1,3,5-tri-terr-butyl-hexahydro-sym-trimine and conformations
in other 1,3,5-triazines
Dehydrosulfurization of Urotropin
Reaction of Carbenes and Carbene Analogs with Alkoxides
and Alcohols
Reactions of Carbenes with Alkoxides and Alcohols
Reactions of L'Si: with Alkoxides and Alcohols
Reactions of LGe: with Alkoxides and Alcohols
Synthesis of 1-Hz
Reaction of 1 with Fe(C0)s
Reaction of 2 with Fe(C0)s
Aromatic Delocalization in Stable Carbenes: Correlation of
Experimental and Computational Data
Introduction
Vibrational Data as Criterion for Aromaticity
Delocalization of Carbene Derivatives
HOMO-LUMO Gaps
Structural Investigation of Carbenes and Protonated Carbenes
The Basicity of Diaminocarbenes
-
Chapter 3
3 .O
3.1
3.2
3.3
3.4
Carbenium Cations as Ionic Liquids
Conclusions and Future Goals
Experimentd 86
General Experimental 87
Synthesis of 2=!3 88
Synthesis of l=S via [1-H] CI 9 1
Attempted Synthesis of 3=S 92
Dehydrosulfurization of 1,3,5-tn-terf-buty l-hexahydro-sym-
95
triazine
Synthesis of 13-Hl SCN
Synthesis of [1-H] Cl
Proton Transfer Reactions (NMR Scale)
[1-H] Cl with D20
[LH] Cl with 2
[2-H] SCN with 1
[3-Hj SCN with 1
[l-Kj Cl with [3-HJ SCN
Synthesis of 1
Synthesis of 2 from 12-81 SCN
Attempted Synthesis of 3
S ynthesis of 1,3,5-tn- te^-butyl-hexahydro-sym-triazine
Synthesis of 1-Hz
Preparation of tert-butoxy lithium
Attempted Synthesis of L'Si-(0tBu)Li
-
Appendix 1
Appendix 2
Appendix 3
Attempted Synthesis of LGe-(0tBu)Li
Attempted Synthesis of 1-(OtBu)Li
Attempted Synthesis of 2-(OtBu)Li
Attempted Synthesis of L'Si-(0tBu)CI
Attempted Synthesis of L1Si(0tBu)2 with BuOLi
Attempted Synthesis of L'S~(O~BU)~ with 'BUOH
Attempted Synthesis of L'Si(0Me)z
Attempted Synthesis of L'Si-(0tBu)H
Attempted Synthesis of LGe-(0tBu)H
Attempted Synthesis of 1-(0'Bu)H
Attempted Synthesis of 2-(0tBu)H
Synthesis of 2-(0Me)H
Preparation of tert-butoxy copper
Attempted Synthesis of L'Si-(OtBu)Cu
Attempted Synthesis of 1=Fe(C0)2
Attempted S ynthesis of 2=Fe(C0)2
X-ray Crystal Structure Data
References
A bbreviations of Compounds
-
List of Tables
Table 1 Table 2 Table 3 Table 4 Table 5 Table 6 Table 7 Table
8
Table 9
Table 10 Table 11 Table 12 Table 13 Table 14 Table 15
Table 16
Table 17 Table 18
Table 19
Table 20
Table 21 Table 22 Table 23 Table 24 Table 25 Table 26
Thioureas in medicine Influence of different reaction conditions
on 2=S, 12-HJ SCN fomation Influence of different reactions on 3-C,
3=S, [3-Hl SCN formation Influence of counter ion and aromaticity
on chernical shifts
% s character in C-H carbon of carbenium salts Solubilities of
carbenium salts Sublimation fractions for deprotonation of [2-H]
SCN with LDA Sublimation fractions for dehydrosulfurization of
1,3,5-tri-tert-butyl-
hexahydro-sym-triazine ( 140 OC, 26 h) Sublimation fractions for
large scale dehydrosulfurization of 1,3,5-tri-te+
butyl-hexahydro-sym-triazine(150 OC,58h) Conformations of
hexahydro-sym-triazines Summary of reactions of 1,2, L'Si:, LGe:
with alkoxides and alcohols Increasing delocalization as obtained
from computational IR frequencies Correlation between bond order
and NMR data Correlation between Eg and aromatic stability
Correlation between computational IR frequencies and Eg for
carbenes and it's derivatives Normal modes, Eg, bond orders, and
experimental NMR data of selected diazoles Normal modes of selected
1,l-dihydro- 1,3-diazoles Comparison of experimental and calculated
structures of L'CH+ cations using B3LYPl6-3 IG*, W/6-3 lG*, MW6-3
le*, and AM 1 methods Comparison of experimental and calculated
structures of LCH+ cations
using B3LYP16-3 L G*, HF/6-3 le*, MP2/6-3 1 G*, and AM 1 methods
Comparison of experimental and calculated structures of 1 cations
using B3LYP/6-3 lG*y HFl6-3 lG*, MP2f6-3 lG*, and AM1 methods
Calculated tme energies &cal) for carbenes and derivatives
Sublimation fractions for Method A Sublimation fractions for Method
B Sublimation fractions for Method C Sublimation fractions for 1-C
formation Sublimation fractions for 3-H2:S8 (1 : 1/4)
vii
-
List of Figures
Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7
Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13 Figure 14
Figure 15 Figure 16 Figure 17 Fipre 18 Figure 19 Figure 20 Figure
21a Figure 21b Fipre 22 Figure 23 Figure 24a Figure 24b Figure 25
Figure 26a Figure 26b Figure 27 Figure 28 Figure 29 Figure 30
Figure 31
Hydrolysis of chlorofom
Representation of electronic structure of carbenes
Arduengo's carbene
Aromatic divalent carbenes la and lb and non-aromatic carbene 2
Synthetic strategies to obtain diaminocarbenes via a carbenoid
Reaction scheme for silylenoid species formation Synthetic
strategy for a-haloorganolithium species
Reaction scheme for carbenoid synthesis
First transition metalcarbene complex
Geometrical positions of the carbene ligand Attempted synthesis
of l=Fe(C0)2 General equation for the synthesis of carbene from
thiourea Literature methods to obtain thiourea General reaction
scheme for the synthesis of thioureas from arninals
Synthesis of the first thiourea Tautomerism in thioureas
Synthesis of cyclic thioureas Thiourea derivatives as
vulcanization accelerators
Synthesis of 2=S from CS2/pylI2 method ORTEP view of [2-Hl SCN
Possible mechanisms for the oxidation of arninals Possible reaction
schemes I,II,III for the sulfurization of aminals Synthesis of 2=S
from dehydrosulfurization of 2-Hz with Se Synthesis of l=S from 1
in a one-pot reaction Possible sulfur containing heterocyclic
compunds Possible structure of the 100-125 O C fraction:
"Zwitterion" formation
Synthesis of 3=S 1H NMR of sublimate at 150 O C (3=S) 1 3 ~ NMR
of sublimate at 150 OC (3=S) ORTEP view of 3-C ORTEP view of [3-II]
SCN Synthesis of [l-8] CI from glyoxal Attempted synthesis of
[l-tI] CI fiom diazadiene ORTEP view of [l-HJ CI
-
Fipre 32 Figure 33 Figure 34 Figure 35 Figure 36 Figure 37 Fipre
38 Figure 39 Figure 40 Figure 41 Figure 42 Figure 43 Figure 44
Fipre 45 Fipre 46 Fipre 47 Figure 48 Figure 49 Figure 50a
Fipre 5Ob Figure SOC Figure 51 Figure 52 Figure 53 Figure 54
Figure 55 Figure 56 Figure 57 Figure 58 Figure 59 Figure 60 Figure
61 Figure 62 Figure 63 Figure 64
Synthesis of 1 by deprotonation of [1-Hl CI ORTEP view of 1
Deuterium exchange reaction of [1-A] CI Proton exchange between
[1-Hl CI and 2 Proton exchange between (2-Hl SCN and 1 Proton
exchange behueen [3-Hl SCN and 1 Proton exchange between [l-Hl CI
and [3-H] SCN Deprotonation of [2-H] SCN to give 2 Deprotonation of
[3-A] SCN to give 3 1H NMR of 80-90 O C fraction (3) Structure of
3-CHO Reaction of SCN- with nBuLi Reaction schemes for [3-H] CI
formation Delocal ized pol y -car benes Synthetic strategies for
tris-carbene Synthesis of Il-Me Decomposition products of the
reaction of 9 with S8 ORTEP view of [l-Hl SCN Reaction scheme for
the dehydrosulfurization of 1,3,5-tri-tert-butyl-
hexahydro-sym-triazine Possible sulfur-exchanp reaction between
9 and 15 13C NMR simulation of mono-, bis-, tri- substituted
thiourea Attempted synthesis of 16 Synthesis of 1,3,5-tn-tert-buty
l-hexahydro-syrn-triazine Graphical representation of melting point
of 9
Synthesis of methylurotropinium thiocyanate
ORTEP view of [5-CH31 SCN Decomposition of halogen-substituted
carbene species to carbenoid Reductive elirnination of aicohols
Synthesis of 2 - ( 0 t ~ u ) ~ i Reaction of 1 with tBuOLi
Reactions of 1,2, LGe: with tBuOH Synthesis of 2-(0Me)H using 1: 1
ratio of 2 to MeOH ORTEP view of 24330 Attempted synthesis of
L'Si-(0tBu)Cu Attempted synthesis of L'Si-(0tBu)Li
-
Figure 65 Figure 66 Figure 67 Figure 68 Figure 69 Figure 70
Figure 71 Figure 72 Figure 73a Figure 73b Figure 73c Figure 74
Figure 75
Attempted synthesis of L'Si-(O*Bu)CI Attempted synthesis of
L'Si-(0'Bu)H Attempted synthesis of LGe-(O%)Li Synthesis of 1-H2
Synthesis of a bis-carbene complex q 1 complex formation: l=Fe(CO)4
Attempted synthesis of 2=Fe(CO)4 The stable carbene 1 and it's
derivatives Reaction between L'CH+ and LC: Reaction between L'CH2
and LC: Reaction between L'CH2 and L CH+ General representation of
ionic liquids Ionic liquids and Carbenium Salts
-
THF Et20 HMPA TMS Me nPr fBu Et iPr PY GCMS IR NMR h d A equiv
mm01 Ph Et3N PES CDC13 C6D6 D2O PP=' r. t. t~ mins. Ca.
SYm asym
tetrafiy drofuran diethyl ether hexarnethylphosphoric triamide
teuamethy lsilane methy l n-propy i tert- buty 1 ethy 1 iso-prop y
l pyridine gas chromatography 1 mass spectrometry in fra-red
nuclear magnetic resonance hour(s) day (s) heat equivalent miili
mole(s) pheny 1 triethy lamine photo electron spectroscopy
deuterated chlorofonn deuterated benzene deuterated water parts per
million room temperature retention tirne minutes approximatel y
symmetnc asymmetric
-
Acknowledgments
1 would like to begin by thanking my supervisor, Prof. Michael
K. Denk, for his
continuous support, guidance and encouragement over the past
twenty months and for giving me
the opportunity to work on some really stimulating projects. 1
like to thank rny CO-workers, past
and present (Ken, Sébastien and Jose), for making the lab a
pleasant place to work in. Thanks aiso
goes to al1 the volunteers in the lab, especially, John and
Neeti, for keeping up with me when 1 got
so very frustrated.
1 am grateful to Dr. Alan Lough for his keen interest and
patience in obtaining those
beautiful X-ray structures. My sincere gratitude gws to the lab
technicians, Sarnia and Pam, for
al1 those srniles and for being there for me. Thanks also goes
to Dr. Tim Burrow for his advise on
some NMR experiments, and to Dan Mathers and Dr. Alex Young for
mass-spectral work.
Lastly, but definiteiy, not the least, 1 like to thank my
parents who have k e n my backbone
every single moment. Words fail to express how deeply grateful I
am for the unconditional love
and moral support they have given me every single day.
xii
-
Chapter 1: Introduction
-
1.1 Carbenes and Carbeniwn Ions
Carbenes are highly reactive species (lifetimes under 1 sec);
the parent species being
CH2 (methylene). The concept that carbenes might play a
significant role as reactive
intermediates dates back to the early kinetic investigations of
Hine who postulated the
intermediacy of C12C: in the hydrolysis of chloroform (Fig. 1) [
11.
- OH . fast
CCI2 - CO + HCO; H20
Fig. 1. Hydrolysis of chlorofon
Methylene has been the subject of the now classical
spectroscopic studies by Herzberg's
group [2]. Unlike most other carbenes, methylene has a triplet
ground state, but the singlet state
is only slightly higher in energy.
The diaminocarbenes la and 2 al1 have a singlet ground state
(Fig. 2). The singlet state
of carbenes with it's empty p-orbital is isoelectronic with
carbocations and is stabilized more by
conjugation than the triplet state which has a singly occupied
p-orbital [3].
Triplet state Singlet state
Fig. 2. Representation of electronic structure of carbcnes
A number of carbenes have been isolated in frozen matrices and
investigated
spectroscopically [4]. Carbenes have been the subject of many
computational studies. Most of
the older results are now obsolete as a result of better and
better experimental data and new
computational data [4]. The extreme reactivity of carbenes made
the goal of obtaining stable
-
15
carbenes seem futile, although early studies by Wanzlick claimed
that diaminocarbenes can
persist in solution and at rwm temperature for prolonged periods
of time.
In 1964, Wanzlick proposed that enetetramines can dissociate
into diaminocarbenes but
was unable to present unambiguous proof for the presence of free
carbenes 151. The goal of
obtaining stable carbenes was finally realized in 1991 when
Arduengo et al. described the
synthesis and structure of
1,3-Di-adarnantyl-irnidazole-2-ylidene, the first unambiguously
stable
carbene [6a-il.
Arduengo's carbene
1 ad
Fig. 3. Arduengo's carbene
Many other studies have since appeared on the subject of
diaminocarbenes from
Arduengo's group [6]. The question if diaminocarbenes require
steric and electronic stabilization
or just electronic stabilization was answered by Our study of
the corresponding saturated
systems. [7].
This study addresses a number of controversial or unexplored
aspects of the chemistry of
stable diaminocarbenes. This thesis c m be grouped around the
subjects:
Synthesis
Aromatic defocalization
Reactivity
New Topologies
-
16
1.1.0 Synthesis
Arduengo's published procedure for the synthesis of stable
carbenes requires the
deprotonation of irnidazolium salts with NaH in DMSO 161. This
rnethod is inconvenient for the
large scale reaction and is aiso unsuitable for the synthesis of
volatile carbenes because of the
separation from DMSO. The use of DMSO is thus problematic for a
general and a specific
reason:
DMSO is not easy to purify and very hygroscopic.
For volatile carbenes, such as 1 and 2 (Fig. 4), the removal of
a high boiling solvent like
DMSO (bp = 189 OC) will inevitably lead to the loss of much of
the formed carbene.
As demonstrated in this thesis, the deprotonation of the
imidazolium salt [l-Hl Cl with
"BuLi in THF is a simple alternative. Separation of the carbene
1 from LiCl formed in the
reaction was no problem and the carbene is easily isolated by
sublimation in Ca. 72 % isolated
y ield.
R R R I I I (2.: R = adamantyl cN): - @,E: mesityl N N
I met hyl
I R R
I R
iso-propyl t e s bu ty 1
1 a l b 2
Fig. 4. Aromatic divalent carbenes l a and l b and non-aromatic
carbene 2
The synthesis of the imidazolium sdts was achieved in a patented
three-component
condensation from glyoxal, a primary amine and formaldehyde. The
synthetic details are
sketchy and the reactions as described are laborious. No proper
purification and work up
procedures are given. This thesis descnbes a simplified
procedure for the synthesis of the
imidazolium salts and two new methods for obtaining the
thiocyanate salts.
-
17
The fact that diaminocarbenes can be obtained by deprotonation
of imidazolium salts
raises the question of how basic diaminocarbenes reaily are.
This question was investigated by
computational methods and by the study of proton exchange
equilibria. The deprotonation of
imidazolium salts with "BuLi and other organometallic bases can
take place by two different
pathways.
Fig. 5. Synthetic strategies to obtain diaminocarbenes via a
carbenoid
These are, general deprotonation that would lead directly to the
carbene (route a) or
metallation that would imply the intermediacy of a carbenoid
(route b) (Fig. 5).
Carbenoids were postulated as intermediates by Witîig et. al. in
1941. According to G. L.
Closs and R. A. Moss, a carbenoid is a species that is
responsible for electrophilic reactions
instead of a carbene (81. The terni is used pnmarily to
characterize a type of mechanistic
behavior and compounds in general that have a metal atom and an
electronegative leaving group
on the sarne carbon atom.
-
1.2 Oxidative Adclition of Alkoxides and AlcohoIs
Analogous to carbenoids, (alkoxysilyl)lithium compounds having
alkoxy groups have
been described by the group of K. Tamao (Fig. 6) [9].
@ \ ,Nu Nu - Si, ' Li Fig. 6. Reaction scheme for silylenoid
species
The possible stability of carbenoids N2CLi-OR and N2CCu-OR was
investigated by
reacting carbenes N2C: with Li-Alkoxides and Cu-Alkoxides as
well as alcohols. Silylenes and
germylenes were likewise investigated. Reactions of carbenes 1
and 2. and germylene, LGe:
with tBu0Li (Fig. 7) and 'BuOCu failed to give the corresponding
alkoxy species excepting
silylene, L'Si:. In the case of the addition of alcohols,
addition was observed for L'Si:, and
carbene 2.
'Bu 1
'BU I
N I
E = C, Si, Ge
Fig. 7. Synthetic strategy for a-haloorganolitùium species
a-haloorganolithium compounds 1.2A are themally unstable (Fig.
8) [ 1 O]. They are
reactive since the heteroatom works as a leaving group.
-
1.2A
Fig. 8. Reaction scheme for carbenoid synthesis
1.3 Metnl-Carbene Complexes
The first synthesis of a transition metalsarbene complex 1.3A
(Fig. 9) by E. O. Fischer
and A. Maasbol in 1964 opened the gates of organometallic
research, such as, in organic
syntheses and catalytic reactions [ 1 I l .
Fig. 9. First transition metal carbene complex
In ail (C0)4Fe(carbene) complexes whose structures are known
from diffraction studies
or spectroscopy, the carbene ligands are good donors but poor
s-acceptors because of one or
two a-substituents having lone pairs. The carbene ligand always
occupies the apical position
(1.3B and 13C), with it's orientation determined by stenc
factors (Fig. 10). This is in accord . with a theoretical andysis
on site preferences for transition metal penta-coordination [ I l ]
.
1.3B 1.3C
Fig. 10 Geometrical positions of the carbene ligand
In this thesis, the reaction of the diaminocarbenes 1 and 2 with
Fe(C0)s was
investigated. The X-ray structure could not be obtained as the
product (yellow, powdery) did not
-
20
crystallize even after layering or by heating the sublimed
product under reduced pressure. The
presence of a carbene complex was concluded from NMR (lH, 1 3 ~
) and Fî-IR spectroscopy.
Solubility problems and synthetic aspects are discussed.
ncat
?
1 1 =Fe(CO)2
Fig. 1 1. Attempted synthesis of 1 =Fe(CO).L
1.4 Thioureas and Thiourea Derivatives
Thioureas are used in the pharmaceutical sector, in plant
protection, in various technical
applications, and in the synthesis of heterocycles [12]. In the
context of this study, we were
interested in thioureas as starting materials for the synthesis
of stable carbenes.
Fig. 12. General equation for the synthesis of carbene h m
thioureû
The synthesis of the parent compound, thiourea NHz-C(S)-NH~ from
calcium
cyanamide is straightforward and is the bais of the current
technical process (SKW, Germany).
A number of different methods are available for the synthesis of
thioureas bearing substituents
on nitrogen (361. Thiourea is not a good starting material for
the synthesis of its N-substituted
derivatives because electrophiles react with the sulfur in most
cases. A closer examination of the
synthetic repertoire reveals serious deficiencies:
-
21
The methods described in the literature are generally
incompatible with bulky substituents
on the nitrogen atoms. For example, compound 4 (Fig. 13) could
not be prepared by known
methods in yields higher than 1
yield - 70 60 50 18 a
a) yield without 12 = O %
Fig. 13. Literature methods to obtain ihiourea
Existing methods typically use the highly toxic and extremely
flammable carbon disulfide or
the toxic and expensive isothiocyanates R-N=C=S as starting
materials.
The standard synthesis of thioureas from isothiocyanates and
amines is restricted to the
synthesis of RI-NH-C(S)-NR*R~.
This thesis presents a new synthesis for diaminocarbenes,
thioureas and stable carbenium
cations. The carbenes are obtained by the reductive
dehydrosulfurization of tetra-substituted
thioureas and I,l-elimination of HCl from carbenium salt,
[N2CH]+ CI-. The thioureas and
carbenium salis are obtained by the subsequent reaction of
amines with formaldehyde and
elemental sulfur. The reaction of the aminals R ~ R ~ N - C H ~
- N R ~ R ~ with typically 114 molar
equivaleni of Se takes place readily between 150 - 180 OC and
leads to the formation of thioureas in modest yields with the major
products being the carbenium saits, [N2CH]+ SCN-.
Fig. 14. General reaction scheme for the synthesis of thioureas
via aminais
-
22
The carbenium cation thiocyanates were converted into the
carbenes. However,
sublimation work up leads to a broad spectnim of decomposition
products (1H NMR).
-
Chapter 2: Results and Discussion
-
24
2.1 Dehydrosulfurization of Aïninais (Rm2CH2 with Sa
Thiourea was first prepared by Reynolds by thermal rearrangement
of ammonium
rhodanide at Ca. 150 OC (Fig. 15) [Ml.
NH4SCN (NHù2CS
Fig. 15. Synthesis o f the first Thiourea
The reaction between carbon disulfide and amrnonia or ammonium
carbonate under
pressure at Ca. 140 O C has not achieved industrial application
[14]. Thiourea has three functional
groups: arnino, imino, and thiol. This results from tautomerism
between thiourea and isothiourea
(Fig. 16).
Thiourea Isothiourea
Fig. 16. Tautomerism in thioureas
Because of this polyfunctionality and also because of its
complex-forming properties,
thiourea has been widely used for more than 30 yean mainly as
starting material for nitrogen-
and sulfur-containing heterocycles and formamidinesulfinic acid,
as a reaction partner for
aldehydes, and as a component of addition cornpounds and
complexes [ 15-19].
Cyclic thioureas have been obtained by the methods known for
open-chah thioureas,
and by the reaction of diamines with thiourea [20].
Fig. 17. Synthesis of cyclic thioureas
-
25
2.1.1 Applications of Thioureas
Thioureas have a wide range of uses, e.g. for producing and
modifying textile and
dyeing auxiliaries [2 1, 221, in the production and modification
of synthetic resins [23], in repro
technology [24], in the production of pharmaceuticals
(sulfathiazoles, tetramisole [25], and
cephalosporins [26], in the production of industriai cleaning
agents (e.g., for photographic tanks
[27], and metal surfaces in general [28. 29]), for engraving
metal surfaces [30], as an
isomerization catalyst in the conversion of maleic to furnaric
acid [3 11, in copper refining
electrolysis [32], in electroplating (e.g., of copper) [35]. and
as an antioxidant (e.g., in
biochemistry) [36]. Other uses are as an additive for slurry
explosives [35], as a viscosity
stabilizer for polymer solutions (e.g., in drilling muds [36])
and as a mobility buffer in
petroleum extraction [37]. The removal of mercury from waste
water of the chlorine-alkali
electrolysis process is possible with thioureas [38]. Thioureas
can also be used to extract gold
and silver from minerals [38,39].
Vulcanization accclcmtor,
H
Fig. 18. Thiourea derivatives as vulcanization accelerators
The N-substituted thioureas, that we are investigating, may find
use as accelerators for
the vulcanization of polychloroprene and
ethylene-propylene-diene terpolymers (EPDM) 1401
(Fig. 18, Table 1).
-
I Compound Use
thyrotherapeutic agent (th yreostatic)
thyrotherapeutic agent (thyreostatic)
thyrotherapeutic agent (thyreostatic)
ultrashort general narcotic, anesthetic
propy lthiouraci 1 :
rnethylthiouracil
thiamy ta1 a
Table 1 Thioureas in medicine
2.1.2 Objectives
Thiourea 2=S was initially obtained in our group from carbon
disulfide / 12 / pyridine:
Fig. 19. Synthesis of 2 5 h m CS2 1 py 1 12 method
-
27
However, this method suffen from the use of toxic and expensive
pyridine and from low
yields (15-20 %). The objective is to directly convert aminals
into thioureas using the one-step
dehydrosulfurization approach . This approach has the advantage
that aminals form readily and in high yields even with sterically
bulky secondary amines, especially if the reaction leads to
cyclic products. The reaction was investigated first for the
Bu-arninal 2-H2 with elemental
sulfur because the desired product 2=S had already been obtained
and hlly characterized from
2-C and CSz. Reaction of 2-Hz with S8 (no solvent) starts at 170
OC as evidenced by the color
change and gas evolution. The reaction gives only a small yield
of 2=S. Investigation of the
sublimation residue showed that the main product (21%) of the
sulfurization of 2-H2 is the
carbenium salt, (2-rn SCN. [&Hl SCN was characterized by
single crystal X-ray diffraction
[data set, appendix I l (Fig. 20).
Fig. 20. ORTEP view with hydrogen atoms omitted for clarity.
Thermal ellipsoids are at the 50%
probability IeveI. Seiected bond distances [pm] and bond angles
[O] as follows: C(1)-N(1) 131.3(2),
C(1)-N(2) 131.4(2), N ( l W ( 2 ) 147.3(2), C(S)-C(3) 15I.7(3),
N(2>-C(4) 149.2(2),
N(3)-C(12) 157.8, N(3)-H( IA) 4 14.6, C(2)-C(3) 15 1.7(3),
N(3-(12)-H(IA) 14 1.43(O.S),
N( I)-C(I>-N(2) 1 13.80(16), C(1)-N(lW(2) 108.83(15),
N(I)-C(2-(3) 102.47(16), H...(N)
24 l(2).
-
28
2.1.3 Mechanisrn and Product Distribution
The breaking of a C-H bond at the comparatively low temperature
of 170 O C is
surprising and requires the proposal of an adequate mechanism.
The strength of a typical C-H
bond mles out a direct homolytic fission. It is likely that the
initial step of the reaction is the
oxidation of the aminal by a S radical (e. g. 'S-(S)6-S) to the
mesomerically stabilized radical
L'CH;?]+' (Fig. Sla).
R-S IL 'BU
I
[;&H
I 'Bu
Fig. 2la. Possible mechanisms for the oxidation of aminals
The radical cation can now loose a proton or a H radical. Both
processes are facilitated
by the fact that the cation and radical character of the
nitrogen atom is partially delocalized into
the C-H bond via hyperconjugation (interaction of the nitrogen
p-orbital with the C-H B*
orbital). The relative importance of the two steps cannot be
evaluated with the data at hand. A
computational study is in progress. For al1 the investigated
reaction conditions, both the thiourea
and the carbenium cation were fonned. Under the reaction
conditions, the carbenium salts and
the thiourea were stable. This rules out the consecutive
formation of one from the other (Fig.
21 b). A branching like the one invoked by the mechanistic
hypothesis above offers a convenient
explanaiion of this expenmental obsetvation .
-
III
Fig. 21 b. Possible reaction schemes 1, II and II1 for the
sulhization of minais.
Apart from a mechanistic insight, the variation of the reaction
conditions had a synthetic
goal, narnely to maxirnize the yield of 2=S or [2-H] SCN. To
this end, three different protocols
(A, B, C) were investigated: the amount of sulfur used is an
obvious parameter to be studied.
2-H2 2-S 2-C (2-H) SCN
Fig. 22. Synthesis of 2=S h m dehydrosulfiirization of 2-Hz with
S8
-
Method A restricts the arnount of sulfur to the
stoichiometrically necessary lower limit
(1/4 equivalent). Method B uses one equivalent which corresponds
to a four fold excess of
sulfur under otherwise identical reaction conditions.
Method A gave the highest yield of 2=!3 (15% ) apart from 12-H]
SCN (2 1 %) and 2-C
(37 96). A mass loss of 1.85 g was unaccountable after adding up
the weights of dl fractions of
the sublimed material.
In order to compare how the different reaction conditions
influence the formation of
2=S, and [2-H] SCN, the weights and % yields of each of the
corresponding fractions were
tabulated (Table 2).
.-
Table 2. Influence of different reaction conditions on 2=S and
[2-Hl SCN formation
The dehydrosulfurization leads to the formation of H2S which in
tum could react with
the arninal by protonation. It was therefore attempted to
increase the overall yield by adding acid
scavengers. However, addition of K2C03 lowered the yields and
gave a more complex product
spectmm as indicated by the 1H NMR of the crude reaction
mixtures. The use of additives was
not pursued any further.
Increasing the amount of sulfur (Method B) leads to a decreased
yield of both the
thiourea and the carbenium salt (10 %). It is noteworthy, that
the thiourea isolated by
sublimation is now contaminated (contrast to Method A) by the
unsaturated thiourea l=S and a
number of other compounds that were characterized only by their
GC-MS traces. The total
combined weight of the thiourea fractions (7.00 g) exceeds the
theoretically possible amount of
5.76 g. It must be concluded, that a substantial part of the
volatile Fraction consists of elemental
-
3 1
sulfur and in fact the total amount of re-isolated sulfur could
be as high as 75 1 (6.16 g) of the
total arnount at the beginning of the reaction.
The yield of [2-H] SCN cm be detemined and allows the conclusion
that an excess of
sulfur is unfavorable for its formation. Method C is identical
to Method A, but the reaction time
has now been increased from 1 h (A) to 30 h (C). This increases
the yield of [2-H] SCN from 21
% (A) to 43 1 (C). This clearly demonstrates. that the formation
of 12-H] SCN is a slow
process that must involve an unknown intermediate. The thiourea
has been ruled out as
intennediate because it is stable under the reaction
conditions.
2.2 l=S via [l-H] Cl
The formation of l=S under the given reaction conditions is the
result of a
dehydrogenation reaction. Pure 2=S can be transformed into l=S
by heating with elemental
sulfur at 170 O C for 18 h. This does not, however, give pure
1=S but always ieads to mixhires of
l=S and the starting material 2=S. Reaction of the stable
carbene 1 with S8 in THF at r. t.
proved to be the only way to obtain pure 1=S (Fig. 23).
1 l=S
Fig. 23 Synthesis of I=S fiom 1 in a one-pot reaction
Reaction of 1 with Sg gave two volatile prducts ( 1 0 - 125 O C
) , l=S and a black micro
crystalline product. This black micro-crystalline compound was
poorly soluble in benzene (5
g/L ) and gave colorless solutions in chlorofomi. This could be
due to decomposition with
CDC13 or poor solubility. The IH NMR in CDClj showed signais at
6(lH, ppm): 1.56 (int. 14 ),
7.06 (d, kt. 1.5 ), 7.36 (int. 2.3), and 7.62 (int. 1).
-
32
These second set of signals can be tentatively ascribed to
compounds of type 2.2A or 23B (Fig.
24a). In view of the volatility of the second component,
structure 2.2A seems more likely.
2.2 A 2.2 B
Fig. 24a. Possible sultùr-containing heterocyclic compounds
The sublimation of the black cornpound between 100 -125 OC
requires that the
compound has a low molecular weight. It is therefore surprising,
that the compound is insoluble
in benzene. The color and solubility of the compound could point
towards a zwitterion Z.
I I 'Bu 'BU
Fig. 24b. Possible structure of the 100-125 OC fraction:
"Zwitterion" formation
2.3 Synthesis of 3=S
The attempts to obtain the thiourea 3-S analogous to 2=S from
the corresponding
diamine 3-C and CS2 (boih with and without the addition of
iodine) led to product mixtures and
insoluble and presumably polymeric materials.
The dehydrosulfurization strategy was tried instead. This
approach was successfbl as
evidenced by the presence of the thiocarbonyl signal (185.3 ppm)
in the I3c NMR spectrum, but large amounts of the 1,3-diamine and
other unidentified / side products were fomed and the
thiourea could not be obtained in pure form.
-
3-Hz 3 5 3-C 13-Hl SCN
Fig. 25. Synthesis of 3=S
The dehydrosulfurization of 3-Hz differs from the analogous
reaction of 2-Hz in three
important ways:
1. The required minimal reaction temperature is substantially
lower for the six-membered
ring (1 10 OC) than for the five-membered ring (160 OC).
2. The amount of diamine formed in the reaction is rnuch higher
in the case of the six-
membered ring than in the case of the five-membered ring.
3. The volatile fraction contains two additional compounds in
substantial quantities for the
six-membered ring, while only traces of impurities were observed
in the case of the five
membered ring.
The reasons for these differences are speculative and further
variation of the ring size
and the steric bulk of the substituents (here: Bu) is clearly
desirable. The formation of the
carbenium cation salt is favored by prolonged reaction times in
both cases.
The number of 'Bu signals (4) in the NMR of the sublimate at 150
O C (Fig. 26a)
suggests the presence of 3=S (1.25 ppm, int. 6.5), J-C (1.12
ppm, int. 1) and two unidentified
side products (1.43 ppm, int. 1.4 and 1.60 ppm, int. 1).
Atso, the 13C NMR of the sublimate shows the C=S resonance at
185.32 ppm which is a
direct evidence for the presence of thiourea (Fig. 26b).
-
Fig. 26a. I H NMR in CDC13 of the sublimate ai 150 O C
Fig. 26b. I3c NMR in CDCl3 otthe sublimate at 150 O C
The yield of 3=S varies between 0% - 40% depending on the
reaction temperature and reaction time (Table 3).
-
3-H2 : S8 3432 : Sa T [OC] crude 3-c 33s [3-H] S a [mols] [ml
t[hl extract 9byields %yields % yields
[ m s l 1:l 1.89 : 2.44 150 3.8gi 2% 40% 122%~
13
Table 3. Influence of different reaction conditions on the
formation of products: X, 3=S, (SHI SCN
this reaction was done in a sublimation f l u k and the crude
mixture was the sum of the weighi of the yellow crystalline solid
on the finger and a black solid at the bottom of the flask. The
residue was a mixture of 3=S, [SHI SCN, 3-C and one other
unidentified product. The weights of the fractions are measured
from the relative intensities of the 'BU signals. ii this fraction
is a yellow oil and the NMR shows a mixture of three different
products with 3=S being the major product.
High reaction temperatures and long reaction times favor the
formation of [3-Hl SCN;
low reaction temperatures favor the formation of 3-C;
intermediate reaction temperatures ( 150
OC) favor the formation of 3=S. Attempts to confirm the
formation of 3=S by an X-ray structure
led to the isolation and structural characterization of the
1,3-diamine 3-C instead (Fig. 27).
Fig. 27. ORTEP view with hydrogen atoms omitted for cliuîty.
Thermal ellipsoids are at the 50%
probability leveI Selected bond distances [pm] a d bond angles
[O] as follows: N(l)-C(l) 146.40(13),
C(1 W ( 2 ) 152.34(14), C(2)-C(3) 152.23(15), N(l)-C(4)
147.88(13), N(1 +C(1 )-C(2) 1 1 1.76(9),
-
36
Attempts to grow crystals of 3=S by sublimation (150- 160 OC) or
crystallization (1 :2
mixture of CHC13 and hexanes) were unsuccessful. The sublimation
residues consist of pure
carbenium sait 13-H] SCN. The sait is obtained in an overall
yield of 34 % (reaction conditions:
190 OC / 40 h) and has been unambiguously characterized by
single crystal X-ray
crystailography (appendix 1, Fig. 28).
Fig. 28. ORTEP view with hydrogen atoms ornitted for clarity.
Thennd ellipsoids are at the 50%
probability level. Selected bond distances [pm] and bond angles
[O] as follows: N(l)-C(I) 130.2(6),
C(2)-C(3) 139.7(7), S( l ) -C(7) 160.0(12), N ( l M ( 4 )
146.6(7), C(1)-N( 1 ) -C(2) 1 16.0(5), N(1)-
C( 1 )-N( 1)#2 129.5( 10). C(3)-C(2)-N(l) 108.9(5), C ( 2 ) # 1
4 ( 3 j C ( 2 ) 24.7(4), H--(N) 394(5).
-
37
2.4 Synthesis of Imidazoüum Sdts from Glyoxai
The imidazolium salts are now of considerable interest as
starting materials for the
synthesis of Arduengo carbenes [6h].They form easily from a
primary amine, glyoxal,
formaldehyde and hydrochloric acid (Fig. 29). The reaction has
so far (1998) only been
descri bed 'BU 'BU
'BUNH~ (2 eq.) ' c P I O 6NHCll24hA THF N
I 'BU
I 'BU
Fig. 29. Synthesis of [l-H] CI fiom glyoxal
in three patents [41a,b] that give no details on the synthesis,
work up and purification. The
mechanism of this one-pot condensation reaction is also
unclear.
The intermediate formation of the corresponding 1,4-diazadienes
from the primary
amine and glyoxal was considered as a mechanistic possibility.
It was therefore desirable to find
out if the imidazolium salts can be obtained directly from the
diazadiene. Mixtures of 1,3-Di-
ter?-butyl-l,4-diazadiene, formaldehyde (35 % in water) and
hydrochlonc acid in different
solvents did not produce any imidazolium salt and only shifts of
diazadiene were present (IH
NMR).
Fig. 30. Attempted synthesis of (1-Hl CI Born diazadiene
It was therefore decided to establish the experimentai details
missing in the patent. The
patent requires first the addition of paraformaldehyde to 6N HCl
followed by addition of tert-
butylamine and finaily glyoxal.
-
38
M e r 24 h stimng at r.t.. the lH-NMR (CDC13) of the aqueous
phase shows the signas
of [1-HICI (1.79, 7.45 and 10.0 ppm) and a second unidentified
component (1.45 s, 4.99 m,
8.30 br., 9.75 br.) The signals of the unknown second component
are in agreement with the
formation of [1-HJOH. An attempt to conven this "hydroxide" into
the chloride with MqSiC1
failed, but the formation of 11-HJCI can be completed by heating
the crude reaction mixture to
reflux for 24 h. The resulting dark brown viscous liquid was
sublimed at 90-180 OC oil bath
temperature. The sublimate is a creamy off-white solid and
poorly soluble in CDC13). it is free
of [1-aC1. The brown. solid sublimation residue is pure [1-HICI.
Total yield is 64 96.
[l-HJCI. was charactetized by single crystal X-ray
crystallography (appendix 1, Fig. 3 1).
Fig. 31. ORTEP view with hydrogen atoms omitted for clarity.
Thermal ellipsoids are at the 50%
probabilîty level. Selected bond distances [pm] and bond angles
[O] as follows: C(l)-Cl(I) 328.4(6),
H(lA)-CI(l) 267(5), C(1j-N(1) 1335(7), N ( l j C ( 2 ) 137.7(7),
N(2)-C(8) 15 1.7(6), C(3)-C(2)
135.9(7), C(1)-N(IW(2) 108.3(4), C(3)-C(2)-N(l) 107.1(5),
N(l)-C(l)-N(2) 108.8(5).
-
39
2.4.1 Deprotonation of [l-Hl CI to give 1
The irnidazolium salt was deprotonated with "BuLi in THE
sublimation gave 72% of
carbene 1 (white powdery, sublimation temp. 60-80 OC).
'BU 'BU I "BuLi THF 1 2S°C I 8 h
b - QU-H (g)
Fig. 32. Synthesis of 1 by deprotonation of (1-Hl CI
Notes:
1. The 1H NMR shifts of the imidazolium salt (in CDCl3) are
critically dependent on the
water content of the compound.
2. The 'Bu-cornpound has not been described in the literature,
but the 1H NMR data of the
ipr-cornPound has been reported and closely matches that of the
Bu compound.
3. Formation of the imidazolium salt is incomplete after 1 h of
reflux.
4. 1H NMR of the crude reaction mixture showed two different
BU-signals.
5. 1H NMR showed the sublimation residue to be pure imidazolium
salt, the sublimate
shows signals at &lH, CDC13, ppm): 1.45 + 1.49 (int. ratio
3: l ) , 1.72, 2.59 ( t ) , 2.77 (d), 4.19 (s, weak), 5.80 (s,
weak) and was not further analyzed.
6. The signals of the sublimate are identical to the impurity in
the crude material.
7. Attempted deprotonation of 11-Ii] CI with THF I NaH at room
temperature did not lead
to the formation of carbene and showed only signals for Il-IQ
Cl.
The carbene 1 was further characterized by single crystal X-ray
crystallography
(appendix 1, Fig. 33).
-
Fig, 33. ORTEP view with hydrogen atoms omitted for clarity.
Thermal ellipsoids are at the 50%
probability level. Selected bond distances [pm] and bond angles
[O] as follows: N(l)-C(I) 136.6(2),
N(l)-C(2) 138.0(2), C(2)-C(3) i34.1(2), N(l)-C(4) 148.9(2),
C(1)-N(lbC(2) 112.57(12),
C(3)- C(2)-N(1) 106.23(14), N(2)-C(1)-N(t) 102.19(12).
2.5 Properties of Imidazolium Salts
2.5.1 Basicity of Carbenes and Acidity of Imidazolium Salts
The imidazoliurn sait [l-ZT] Cl is the C-protonated derivative
of the stable carbene 1. It
seemed interesting to compare the structures of the carbene with
the structure of the carbenium
cation and to establish the relative basicity of different
carbenes through proton exchange
reactions. In a preliminary snidy, it was investigated if the
imidazolium salt possesses any
significant CH acidity . To this end, mixture of [l-8] Cl and
D20 was investigated, but did not
show any H / D exchange products.
-
Fig. 34. Deuterium exchange reaction of Il-Hl CI
It can therefore be concluded, that the acidity of the
imidazolium salt is quite low,
presumably c 20. The [1-Hl Cl salt was first pumped Ln vacuo at
oil bath temperature of 180-
190 O C for 18 h to dry the salt. The exchange experiment was
perforxned in a flame sealed NMR
tube. No exchange was observed over a period of 7 d at 25 OC and
at 1 10 O C for 22 h.
The relative basicity of the carbenes 1 and 2 was studied
through a competition experiment.
Fig. 35. Proton exchange reaction behveen Il-Hl Ci and 2
A sealed NMR sample of an equimolar mixture of [bH] Cl and 2
(Fig. 35) in C a 6
showed signals for the conjugate base, narnely the carbene 1.
The two carbenes are present in a
ratio of 1 1 2 = 1 1 18. The signals of the carbenium salts were
not observed because they are
insoluble in C&j.
For the mixture [2-8] SCN I l (Fig.36), equilibration led to a
ratio of 1 1 2 = 7 1 1.
-
'BU 'BU 'BU 'BU 1 SCN- 1 1 SCN - 1
CbDb [Y. + cN): - N 25 O C ["H N + [)
I 'BU
I 'BU I 'BU I 'BU
12-Hl SCN 1 [l-HI SCN 2
Fig. 36. Proton exchange behveen I2-HI SCN and 1
After 21 d, this ratio changed to 1 I 2 = 2.7 1 1 (Fig. 36).
This indicates that the
protonation equilibrium is slow.
Apart from the tBu signals of the newly formed carbeme 2 (1.36
ppm). new signals at
1.67. 1.83, 2.73, and 6.29 ppm which contribute to the formation
of [l-H] SCN (1.83, 6.29). It
is noteworthy, that the saturated carbenium salt [2-Hl SCN is
insoluble in benzene (absence of
signals in C6&) while the unsaturated salt [l-H] SCN is
soluble in benzene (20 gL) . Although the data indicate that the
two carbenes are of similar basicity, the obtained data
could also reflect the relative solubilities of the two
imidazolium salts. It was therefore
necessary to repeat the investigation in a solvent that
dissolves d l 4 compounds without reacting
with them. THF was investigated but found unsuitable because of
signal overlap with the b u -
signals and the N-CH2 signals. The 13c NMR of the mixture of 2
and [1-Hl CI (Fig. 35) showed resonance at 16 1.46 ppm which
indicates the formation of [ 2 - 9 CI. The signals for Ç(CH3)3
and CH3 were hidden under THE So, a different solvent was
desired that would dissolve both
the reactants and the product.
HMPA dissolves [3-R] SCN but does not dissolve [1-A] CI.
Although HMPA is very
inert towards reducing agents - many reductions with elemental
Li, Na and K can be
conducted in HMPA - oxygen transfer reaction between HMPA and
the carbenes 1 and 2 can not be ruled out. The sarne reaction was
tried in HMPA, but 11-HJ CI was completely insoluble
in HMPA even though 2 is soluble and stable in HMPA. The
reaction mixture only showed
chernical shifts for 2.
-
43
It was found, that the unsaturated carbene 1 is inert towards
HMPA. A mixture of 1 and
HMPA (1: 1) in C& showed only carbene signals at 1.44 and
7.09 ppm. In neat HMPA the
shifts are 1.54 and 7.34 ppm. These values are slightly shifted
vs. the values determined in
C@6. The signal of the ring protons is slightly broadened. The
reason for this line broadening is
unclear.
A mixture of 1 and [3-H] SCN in HMPA was also measured with a
D20 insert and TMS
as intemal reference (Fig. 37). Although the signals for the
supposedly formed carbene 3 could
not be unambiguously assigned because they are partially hidden
by the strong and broad
HMPA signal at 2.6 ppm. proton exchange must have taken place
because the carbene 1 has
been consumed (no signals) and the salt [l-Hl SCN has been
fonned (1.76,7.2 and 10.7 pprn).
'BU 'Bu 'Bu 'Bu I ' s o l - 1 1 SCN -
HMPA/DtO CF. + cN): N - 25 OC I I
'BU 'BU I
'Bu l
'BU
13-H] SCN 1 3 [i-Hl SCN
Fig. 37. Proton exchange between 13-HI SCN and 1
Due to it's poor solubility in benzene (10 g L ) , the signals
of [3-H] SCN were completely
invisible. The signal at 1.38 pprn is assigned to the new
carbene 3, the signal at 4.5 pprn to
HDO. A signal at 1.02 pprn remains unaccounted for.
The 13~(1H) spectrurn of the sample shows the signals for [1-H]
SCN (29.49
[C(çH3)3], 60.12 [Ç(CH3h], 121.45 [ÇH=ÇH], 134.66 [ç+-Hl, 160.91
[SÇN-1) and for 3
(29.04,39.42,60.78). The signal for the carbene carbon of 3 was
not observed, presumably due
to low intensity.
-
Although the carbenium salts show variation in solubility, they
are al1 non-volatile. Their
ionic composition was verified by X-ray structures (appendix 1).
The structures do noi show any
covalent bonding between the carbenium ions and the counter ions
CI- and SCN-. In solution,
the NMR shifts of the carbenium salts depends on the nature of
the counter ion (Table 4). This is
strong indication for the formation of ion pairs in
solution.
11-a CI Il-H] SCN [2=Ef] Cl [SEI] SCN W(CH3)3) 1.8 1 1.84 1.55
1.52
NCcH) 7.7 1 7.4 1 4.05 4.1 1 & C D 0 119.71 1 19.47 57.20
56.86
Table 4. Influence of counter ion and aromaticity on the
chemical shifts
A mixture of [LH] Cl and [3-H] SCN was investigated to find out
if the contact ion
pairs exchange rapidly on the NMR time scale. At room
temperature, there was only one set of
signals for each of the salts. This implies a rapid exchange of
the counter ions; in the case of
slow exchange, signal broadening or even four different sets of
signals would be expected.
[i- H] Cl [3- H] SCN 11- H] SCN 13- Hl CI
Fig. 38. Proton exchange between 11-m CI and [SHI SCN
-
The acidity of C- H bonds cm be estimated from the ' J (C,H)
coupling constant. The
empirical relation denved from compounds with known
hybridizations like ethylene (sp2) is:
IJ (c,H) = 5 * (% S) [in Hz1
Although the equation [69] is strictly valid only for
hydrocarbons, it can nevertheless be used to
estimate the relative acidity of other sets of closely related
compounds. For the carbenium
cations the following sequence of relative C-H acidities was
established (Table 5) .
carbenium 1 J (c,H) 96s p sa1 t [Hz] [l-Hl CI 2 19.7 44 [2-H]
cli 201.4 40 [l-Hl SCN 201.3 40 12-H] SCN 199.9 40 13-H] SCN 189.1
3 8
Table 5. % s character in the C-H carbon of the carbenium
salts
i obtained by I. Rodezno (unpublished results)
The data are interesting in two respects. First, the effect of
ion pairing is clearly visible
from the difference of the coupling constants for the pair [1-H]
SCN and [l-H] CI (Av = 8.4
Hz). Second, for [2-H] CI and [2-H] SCN, the coupling constants
are very similar (Av = 1.50
Hz). It is also interesting to note that different types of
carbenium ion salts can show similar
CH-acidity, e.g., 12-Hl CI / [1-Hl SCN (Av = O. 10 Hz).
2.5.4 Solubilities of Carbenium Salts
The solubility of the carbenium salts in waier and organic
solvents was investigated to
properly plan the synthetic work-up and reactions. Table 6
reveals the important influence of the
counter ion.
22 50 20 [SHI SCN 100 1 50 10
Table é. Solubilities of carbenium salts in g/L
-
46
While the chloride [1-Hl CI is very soluble in chloroform and
water, the thiocyanate [l-Hl SCN
is only moderately soluble. A possible explanation is a higher
degree of covalency in the
thiocyanate. Solubility data of [2-Hl salts will eventually
complete the picture and allow a
consistent interpretation.
2.6 Deprotonation Stntegy to give 2 from [2-H] SCN
The fact that the thiocyanate salts [1-Hl SCN. 12-Hl SCN and
[3-Hl SCN can be
obtained from very inexpensive starting materials through
hydrodesulfurization makes them
attractive starting materials for the synthesis of stable
carbenes.
'BU 'BU
LDA (2.1 cq.) LiSCN 25 OC, THF
I - LDA-H I 'BU 'BU
(2-HI SCN 2
Fig. 39. Deprotonation of 12-HI SCN to give 2
A number of different deprotonation bases were investigated.
Diaminocarbenes were
fonned, but, upon attempted sublimation, only decomposition
products were isolated. This
cannot be explained with low thermal stability of the carbenes,
as al1 carbenes under
investigation were previously obtained by alternative methods
(deprotonation of the carbenium
chloride salt, reduction of thioureas) and found to be thermally
stable. The characteristic odor of
sulfur compounds noticed in the decomposition products suggests
most likely the reaction of
carbene 2 with the counter ion SCN-• This hypothesis suggests a
possible modification of work
up to prevent the decomposition reaction. Work up in this thesis
consisted of simple transfer of
the crude reaction mixture to the sublimation flask. If
decomposition is indeed caused by the
reaction of carbene with the side product LiSCN, the high
solubility of the carbene in
hydrocarbons should allow separation of the carbene from the
thiocyanate prior to sublimation.
-
47
The deprotonation base and the nature of the carbenium cation
were found to be important.
While [3-Hl SCN gave the conesponding carbene &er
sublimation when LDA in THF was
used as deprotonation base, no carbene could be isolated when
"BuLi was used as base.
In the case of 12-H] SCN, the carbene could be generated with
LDA but decomposes upon
sublimation work up.
Ring opening hydrolysis of the carbenes 2 and 3 was found to be
a side reaction during
the isolation of 2 and 3. The 1H NMR of the sublimate at 50 O C
showed resonances at 6 ( 1 ~ ,
C&, ppm): 1.32 (int. l), 1.36 (2, int. 1) and 1.55 (int.
3.5) along with resonances for 2-CHO
6(lH, CaDa, ppm): 0.85(s), 1 .Ol(s), 2.7 1(t), 3.34(t), and
8.40(s). Signals of 12-H] SCN were
absent. The signais at 1.32 and 1.55 ppm remain to be
identified.
2.7 Deprotonation Strategy to give 3 from [3-H] SCN
The synthesis of 3 from the easily accessible thiocyanate 13-Hl
SCN by deprotonation is
particularly important in view of the inaccessibility of the
thiourea 3=S (Fig. 40).
'Bu 'BU 'BU I
il C - H -
THF NH I
'Bu
[SHI SCN 3 3-CHO
i ) 2. t equiv. of "BuLi or LDA or Cdimethylamino-pyridine
Fig. 40. Deprotonation of [%Hl SCN to give 3
The sublimation gave the following fractions (Table 7).
Sublimation Weigbt Appearanee
90 - 100 0.04 black grandar residue 0.5 1 black solid
Table 7. Sublimation hctîons for the deprotonation of P-m SCN
wiîh LDA
-
48
The coiorless liquid (80-90 OC, Table 7) was identified by 1H
NMR to be 3 (6(1H): 1.38,
1.67,2.67 ppm) (Fig. 41).
n t -
Fig. 41. I H NMR in CgDg of the 80-90 OC fiaction (colorless
oil)
The sublimation residue contained a rninor product with signais
at 0.85 and 1.03
ppm. This product is tentatively identified as the ring opened
hydrolysis product 3-CHO. This
interpretation is also supported by the isolation and structurai
characterization of the analogous
2-CHO.
Fig. 42. Stnicture of 3-CHO
2.8 Other Deprotonation Bases
4-Dimethylamino pyridine as possible deprotonation base was
tested for both [2-H]
SCN and 13-H] SCN but did not react (r. t., 4 h, 1H NMR). The
use of "BuLi was also
unsuccessful in the deprotonation of [2-tn SCN and [3-EI] SCN.
This is in line with the known
nactivity of the thiocyanate ion (Fig. 43).
-
Fig. 43 Reaction of SCN' with "BuLi
DMF Chloride
'BP Mtthod 1 rNH DMF Chloride
Fig. 44. Reaction schemes for 13-HI CI formation
'9 ,& Mtthod 2 *O C h
PbCI2 or HCI N te:
A solution to this problem would be the use of chloride salts
instead of thiocyanate salts.
Two methods for the synthesis of chloride salts are curiently
being investigated (Fig. 44).
-
50
f 9 Aromatic, Anü-Ammatic d Linear Coqjugated Poly-carbenes
The strong aromatic delocalization in the diaminocarbene 1
suggests that polycarbenes
of the general type (R-N-C:), should also be delocalized.
Fig. 45. Delocalized poly-carbenes
Calculations (M. Denk, unpublished results) indicated, that the
aromatic carbene 7 (Fig.
45) is perfectly planar (MP2 / 6-3lG*) (LNCN = 108.20, LCNC = 13
1.80) while the anti-
aromatic carbene 6 adopts an envelope geometry (LNCN = 87.g0,
LCNC = 85.30). The C-N
bond distances in 7 were found to be equal and short (137.6 pm).
The anti-aromatic carbene 6
shows elongated C-N bond distance ( 14 1.3 pm).
None of the representatives in Fig. 45 excluding 1 has been
obtained. Especially, the
aromatically stabilized tris-carbene 7 was considered to be a
candidate of potential high
stability. Possible strategies for the synthesis of 7 are
oudined in Fig. 46. Investigations in this
thesis are restricted to the synthesis from the possible
desulfurization of tris thiourea 11. The
desulfurization method had been previously developed in Our
group and is well suited for the
synthesis of both stable and transient diaminocarbenes [7]. The
trimerkation of isonitriles is
endothermic (MP2 / 6-3 1 G* level).
-
5 1
The synthesis of the tris thiouna 1lfBu proved to be a
formidable task. While the îris-
methyl derivative Il-Me cm be obtained via a traditionai
multistep synthesis. al1 other membea
have k e n obtained from ultra high pressure trimerizations
[42].
w
dchydrogcnation
\ k//" F/ F
RA *\R fi' S
11 only known for R = Me, Et, "Pr, "Bu, Ph
Fig. 46. Synthetic strategies for tris-carbene
The group of Yoichi Taguchi in Japan succeeded in trimerizhg Me
isothiocyanate (12-Me) to
1,3,5-trimethyl- l,3,5-triazine-2,4,6 (lH,3H,SH)-tnthione
(Il-Me) (Fig. 47) under high pressure
(800 MPa) [42]. The reaction also requires catalysis by Et3N: or
pyridine. Sterically hindered
pyridines do not catalyze the reaction.
Fig. 47. Synthesis of 1bMe
-
52
The rate of trimerization of 12-Me was found to be proportional
to the amount of
triethylarnine catalyst used. Ethyl isothiocyanate also
trimerized in high yield under the same
reaction conditions. Bulkier groups like R = "Pr, "Bu, allyl,
Ph, and cyclohexyl gave very low
yields. A second protocol using DBU as catalyst under 800 MPa at
130 O C for 20 h resulted in
the trimerization of "Pr and "Bu isothiocyanates in good yields
but failed for larger alkyl groups
( i ~ r , tBu). Allyl isothiocyanate polymerized in the presence
of DBU but the trimer was obtained
in the presence of triethylamine at even higher pressures ( 1200
MPa) in good yield (421.
This serves to illustrate that general methodologies for the
synthesis of tris-thione-
triazines 11 remain yet to be developed.
Attempts by Martin Ma in Our group to develop catalysts for the
low pressure
trimerization of 12 to give 11 were unsuccesful. It was
therefore decided to investigate the
transformation of 9 -> 11 by dehydrosulfurization. The
transformations 9 -> 7 and 10 -> 7 are
cumntly under investigation in our laboratory. The
transformation of aminals into carbenium
cations is demonstrated in this work for 1-Hz, 2-Hz and
3-H2.
2.10 Reactivity of lJJ- tri-tefi-butyl-hexahydro-sym-triazine
with Sg
The reaction of 1,3,5-tri-tert-butyl-hexahydro-sym-Win with
eIemental sulfur ai 140 -
160 OC gave mixtures that were separateci by sublimation (Fig.
48). The degradation products 13
and 14 that form the volatile part are readily explained. No
mechanistic explanation can be
offered for the formation of [1-HJ SCN (17% yield, sublimation
residue). Although surprising
and unexplained, the formation of [l-H] SCN (17 %) constitutes
an inexpensive and fast
synthesis for this compound.
Sublimation weight Appearance Assignment Temp. [% yield]
--80 35 crystalline, yellow 14 80 - 120 36 crystalline, yellow
13 120 - 140 - oily, orange 13114= 1.5/1 residue 17 flaky, oranp;e
[l-a] SCN
Table 8. Dehydrosul~t ion of
1,3,S-tri-tert-butyl-hexahydr0csym-triaPne (140 OC, 26 h),
-
53
The absence of Il-'Bu was confirmed by NMR (IH, ' 3 ~ ) and
GCIMS. Upon dissolving
the sublimation residue in CHC13 and layering with twice the
volume of hexanes, a brown
precipitate was obtained.
'BU SCN - I [EH + I
'Bu
Il-HI SCN
314 S*
'BU
yN4 , N ~ N \
'BU S 'Bu
Fig. 48 Decomposition products of the reaction of 9 with Sg
This presumably polymeric material is insoluble in water and
displays a characteristic
NH band (3643 cm-'), and a strong and broad band at 2061 cm-1
(-N=C=S), and an N-C=C
band (1651 cm-').
While the formation of 13-Di-teri-butyl thiourea 13 is supported
by NMR (IH, 1 3 ~ ) and
GC/MS, evidence for the formation of tert-butyl thiourea 14 is
less convincing (lH, 1 3 ~ NMR).
Table 8 illustrates the fractions obtained by a typical
work-up.
The CHCl3-soluble part of the sublimation residue is pure [l-Hl
SCN and was
characterized by single crystal X-ray crystallography (Fig.
49).
-
Fig. 49. ORTEP view with hydrogen atoms omitted for clarity.
Thermal ellipsoids are at the 50%
probability b e l . Selected bond distances [pm] and bond angles
[O] as follows: S( 1 )-C( 12) 164.94( 18).
Repeating the reaction led to the product distribution outlined
in Table 9.
sublimation temp. weight ~ppearance - ~orn~chents-. r Oc1
Egramsl
6 0 - 100 1.55 oily, yellow 2 + 14 100 - 140 140 - 180 180 - 195
195 - 200 200 - 220 residue
1 .O0 flaky, yellow 2 O. 12 creamy, yellow 2 0.47 creamy, yellow
2 0.26 oily, yellow [BUNCS + 15 0.63 crearny, pale yellow 2 + 14
0.5 flaky, pale yellow 1 + ~BUNCS +
15
Table 9. Sublimation fiactions obtained fiom large scale
dehydrosulfuritation o f 1,3,5-tri-tert-
butyl-hexahydro-sym-criaine at 150 OC, 58 h
None of these fractions contained Il-tBu and the residue was not
[1-H] SCN (lH, 1 3 ~ ,
GC/MS). A possible reaction scherne for the formation of the
tris-thiourea would be as follows
(Fig. SOa,b) .
-
'BU, 4 'Bu HN N'
Fig. SOa. Reaction scheme for the dehydrosulfurization of
1,3,5-tri-tert-butyi-hexahy&o-sym-triazine
Fig. SOb. Possible sulk-exchange reaction between 9 and 15
-
56
A 13C NMR simulation software was used to simulate t h e I 3 ~
NMR shifts for mono-,
bis- or tri-substituted thioureas (Fig. SOC).
Fig. SOC. I3c NMR simulation of mono-, bis-, tri- substinited
thiourea
2.10.1 Multistep Approaches for the Synthesis of Il-'Bu
Compound 16 is a necessary intermediate in the synthesis of
Il-
-
57
This may be due to the superposition of the signals of 16 and
another unidentified product. The
formation of the triazine 9 c m be ruled out on the basis of its
1H NMR data (1.05, 3.45 ppm).
The 13-Di-tert-butyl thiourea used in the reaction was
contarninated with 28% mono-tert-butyl
thiourea (IH NMR). This presumably led to a mixture of products
upon reacting it with
paraformaldehyde and tert-butyl amine.
The 13C(lH} spctmm of the crude in CDCl3 showed strong signals
at 29.41 and 29.87
ppm, with weak signals at 30.97, 122.63, 128.5, 134.25 and 224
ppm. No signal around 180
pprn was observed and also no signals for quarternary carbon
around 50 ppm were observed.
From the FT-IR data (NaCl, nujol), bands at 3301 and 3266 cm-'
appear as a doublet indicating
an NH2 stretch.
Al1 these results support the absence of the mono-substituted
thiourea and the presence
of a mixture of 13 and 14 (1H NMR).
2.11 1,3,5-tri-tert-butyl-hexahydro-sym-triazine and
conformations in other 1,3,5-triazines
As discussed in chapter 2.9,
1,3,5-tri-tert-butyl-hexahydro-sym-triazine 9 is a potential
starting material for the synthesis of the tris-carbene 7. Like
most hexahydro-sym-triazines, 9 is
in equilibrium with the imine CHZ=N(~BU) 4. This chapter will
discuss the synthesis and
properties of 9 and its derivatives.
1,3,5-triazacyclohexane, the most simple of hexahydro-triazines
was obtained in 1895
from a mixture of aqueous formaldehyde-ammonium chloride upon
addition of potassium
carbonate but has never been obtained as such [42]. Obviously,
this compound transforms into
the very stable urotropin with it's tetra-aza-adamantane
structure. In solution, the compound is in
equilibrium with irnine CH2=NH. Hexahydro-synt-triazines with
substituents on nitrogen can
not form urotropin and are typically more stable, but can also
dissociate into imines.
Hexahydro-sym-triazines can exist in different conformations as
Table IO shows
structurally c harac terized exarnples (Cambridge Database up to
May 1 998).
-
58
As the data show, the equatonal, axial, axial (eaa) conformation
is the most common,
eea and even aaa conformations are known.
Ri = R2 = R3 Conformation Ref. aaa 57 eaa eaa eea eee eee eea
eea eea eaa eaa eaa
PhOCH7- eaa 66
Table 10. Conformations of hexahydro- l,3,5-triazines
(hexahydro-sym-uiazines)
Reaction of 'Bu-NH2 with parafomaldehyde gives
1,3,5-tri-tert-butyl-hexahydro-sym-
triazine after one day of reaction at r. t. The use of
paraformaldehyde is convenient because its
dissolution allows to monitor the reaction [S. Rodezno,
unpublished results]. The triazines
separate from aqueous solution as oily layers that c m be
separated and drkd with Ba0 or KOH.
For MeNH2, no phase separation was observed and the product was
isolated by repeated
extraction with Et20.
Reaction time is only 20 mins. in this case. 33% solutions of
fonnaidehyde react even
faster but lead io products of lower punty.
ta..
Fig. 52. Synthesis of
1,3,5-tri-tert-butyl-hexahydro-sym-tnazine
As generally found for hexahydro-sym-triazines, compound 9 is in
equilibrium with 4.
The formation of 9 is favored in concentrated solutions.
-
59
This equilibrium leads to the pualing observation, that in
C&, the ratio of 9 1 4 is
concentration dependent. As expected, dilution leads to a
decrease of 9 1 4. In the more polar
CDC13, sipals for 9 are weak or even absent, while the polar
imine 4 is obviously stabilized by
polar solvents.
Attempts to obtain crystds for the triazine 9 failed and gave
only sticky oils. This could
be due to the fact that 9 is a mixture of different anomers.
However, a melting curve recorded
for 9 gave only one small inflection point at 5 OC (Fig.
53).
Melting point of 1,3,5-tri-tee-butyl hexa hydro-s- triazine
Fig. 53. Graphical representation of the meiting point of 9
-
60
2.12 Dehydrosulfurization of Urotropin
Urotropin is nlated to the hexahydro-s-triazines studied in this
thesis. Attempts to obtain
dehydrosulfunzation products with cage structure led to the
formation of methylurotropiniurn
thiocyanate [S-CH31 SCN as the only identifiable product (Fig.
54). The compound was
characterized by single crystal X-ray crystallography (Fig.
55).
5 [S-CH3] SCN
Fig. 54. Synthesis of methylurotropiniurn thiocyanate
+ unidentified cornpounds
Fig. 55. ORTEP view with hydrogen atoms omined for clruity.
Thermal ellipsoids are at the 50%
probability level. Selected bond distances [pm] and bond angles
[O] as follows: N(4)-C(6) 116.2(3),
S(I>-C(6) 165.2(2), N(2)-C(2)#1 153.12(17), N(2+C(5)
148.0(3), C(3)#1-N(3) 147.59(17),
C(SbN(2>-C(2) 1 1 l.l4( IO), C(2+N(2+C(l) lO8.l7(lO),
C(2*1-N(2)-C(l) 108.17(10), N(1+
C(2+N(2) 1 10.18( 121, N(l)-C(3&N(3) 1 1 1.78(12),
N(lH(4)-N(1)#1 1 1 1.80(16), N(4)-C(6)-
S(1) 179.7(2).
-
61
The fluoride analog of the salt has been introduced recently as
a source of naked fluoride
ions by Clarke et al. Methylurotropinium fluoride dihydrate can
be obtained from urotropin and
methyl iodide via 1-meihylhexamethylenetetraarnine iodide, which
is then converted to the
corresponding fluoride by metathesis with AgF. The compound has
high thermal stability and
the large size of the cation makes it a g w d source of F- ions.
More recently, Robert Gnann and
coworkers have reported a single-step one pot synthesis of this
compound and it's application in
the isolation of the anhydrous fluoride as well as in a coupling
reaction [43].
2.13 Reaction of Carbones and Carbene Analogs with Alkoxides
The oxidative addition of metal salts MX with X = RO, Hal, NR2
etc. leads to
carbenoids. Carbenoids are typically obtained by lithium-halogen
exchange from 1, l-dihalo-
alkane. They only possess marginal stability and decompose above
-100 O C to give the typicai
decomposition products of free carbenes [44].
Fig. 56. Decomposition of halogen substituted carbene species to
carbenoid
The reductive elimination of alcohols from ester-arninals
N2CH-OR to give carbenes
N2C: has been applied for the synthesis of tetraamino-olefins
from amines and ortho-esters (Fig.
57).
Fig. 57. Reductive elirnination of alcohols
-
62
There is, however, no evidence for the operation of this
mechanism other than the formation of
the enetetraamine. Even the thermodynamic equilibrium between
N2CH-OR / N2C: + ROH is unclear at present. Silylenoids, the
sila-analogs of carbenoids have received little attention but
are now investigated by the group of K. Tamao at Kyoto [44]. The
possible oxidative addition
of alcohols to carbenes 1 and 2, L'Si: and LGe: was investigated
to this end.
The oxidative addition of tBuOCu, 'BuOLi, tBuOH and MeOH to the
carbenes 1 and 2
as well as their sila- and germa-analogs was investigated. The
[BuOLi in this study was obtained
from BuOH (as received) and nBuLi.
The [BuOCu has the unique property of strong affinity towards
n-accepting ligands [93,
941 and was obtained from tBuOLi and CUI and purified by
sublimation ( 130 - 150 OC, 0.1
Torr).
The reaction of tBuOH with carbenes 1 and 2, L'Si: and LGe: was
investigated. To
study the effect of steric hindrance, BuOH was replaced with
MeOH.
2.13.1 Reactions of Carbenes with Alkoxides and Alcohols
Carbenes 1 and 2 did not react with tBuOCu both at r. t. and
after prolonged heating at
110 OC for 7-10 d. The reaction of carbene 2 with tBuOLi showed
absence of starting materiai
and the presence of new signals at 1.32 (2H), 1.35 (1 H) and
3.05 ppm (0.5H) that suggested the
formation of 2-(0t~u)~i
dilute sample.
1 'BUOL~ - THF
(Fig. 58). A 1 3 ~ NMR of this sample could not be obtained due
to a
' ~ p
[NU"" + I
'Bu
Fig. 58. Synthesis of 2-(0fBu)~i
-
63
Attempts to repeat the reaction only led to the formation of the
hydrolysis product 2-
CHO that was charactenzed by single crystal X-ray
crystallography. No reaction occurred
between tBuOLi and 1 (25 OC 1 7 d followed by 100 O C 1 14 d)
(Fig. 59).
'BU 'BU I I
1 'BUOL~ N O 'BU
N Li I
'BU I
'Bu
Fig. 59. Reaction of I with 'BuOLi
Reactions of tBuOH with 1 and 2 only gave hydrolysis reaction
and were not pursued
any further (Fig. 60).
Fig. 60. Reactions of I,2 and LGe: with 'BUOH (1:2)
No reaction was observed between 1 and MeOH even after heating
over a penod of 7 d
ai 100 O C . Reaction of 2 with MeOH showed signals for 2-(0Me)H
at &'H, C&, ppm): 1.13,
2.73, 2.94, 3.24, 5.23 and a minor product: 6(lH, C6D6, ppm):
0.86, 1.01, 2.57, 8.40 identified
as 2-CHO upon sublimation work up (Fig. 61).
-
'BU 1 McOH 'BU I 'Bu
I 25 OC
Mat
I NH
'BU I
'BU I
'Bu
2 240Me)H 2-CHO
Fig. 61. Synthesis of 24OMe)H using I : 1 ratio of 2 to MeOH
Compound 2-CHO was characterized by single crystal X-ray
crystallography (Fig. 62).
Fig. 62. ORTEP view with hydrogen atoms ornitted for clarity.
Thermai ellipsoiâs are at the 50%
probability level. Selected bond distances [pm] and bond angles
[O] as iollows: N( l ) -C( l ) 134.14(18),
N(l s ( 2 ) l49.38( 18). C( 1 ) -C(6) 152.5(2), C(6)-N(2)
145.86(18), O(I)-C(11 j N ( 1 ) l23.98(14),
C(t)-N(lH(2) 1 l8.7O(ll), N( l ) -C( l jC(6) II4.15(12).
-
2.13.2 Reactions of L'Si: with Alkoxides and Alcohob
The reaction of L'Si: with tBuOCu gave new product signals at
&lH, C6D6): 1.28
(18H), 1.32 (9.8H), and a broad doublet at 5.80 (PH) after
heating (LOO OC, 10 d) dong with a
color change of the solution from orange to dark brown and a
black film on the upper parts of
the NMR tube. This indicates the formation of copper requiring
the oxidation of L'Si: (1H and
13C NMR) (Fig. 63).
L'SI: L~S~-(O~BU)CU
Fig. 63. Attempted synthesis o f L's~-(O
-
66
Attempts to obtain compounds of type L'Si(0R)Cl or L'Si(0R)z
from the silane L'Sic12
did not succeed. Reaction of L'Sic12 with tBuOLi (1:l) at 25 O C
or at refluxing temperature
produced a color change but no new products (Fig. 65). Use of
BuOH + pyridine instead of BuOLi also did not give substitution
products.
8 THF
N CI I
- LiCl 'BU
I 'BU
L'Sic12 L'S~-(O~BU)CI
Fig. 65. Attempted synthesis of L' s ~ - ( O ~ ) C I
tBuOH readily adds to the silylene L'Si:. The product was
isolated by sublimation of the
crude mixture 45 O C - 80 O C (oil bath, 0.1 Torr) as off-white
crystals (56% yield) and
characterized as L'Si-(0tBu)H (IH, 13C NMR) (Fig. 66).
'BU 'BU l I
2 'BUOH
N THF H I
'BU I
'BU
L'Si: L's~- (0%)~
Fig. 66. Attempted synthesis of L~s~-(oBu)H
Surprisingly, no reaction of L'Si: was observed with MeOH.
-
2.13.3 Reactions of LGe: with Alkoxides and Alcohols
No reaction was observed between LGe: and tBuOCu or tBuOLi (Fig.
67).
LGe: LG~-(O'BU)L~
Fig. 67. Attempted synthesis of LG~-(O~BU)L~
Reaction of LGe: with 'BUOH (Fig. 60) or MeOH only led to the
formation of 2-C and
the starting material LGe: ('H NMR). Table 11 summarizes the
outcome of the reactions of
carbenes 1 and 2, L'Si: and LGe: with alkoxides and
alcohols.
t ~ u ~ ~ u t ~ u ~ ~ i t ~ u ~ ~ MeOH
1 - - - - 2 - + + +
L'Si: + + + - LGe: - - + +
Table 11. Surnmary of reactions of 1,2, L'Si:, LGe: with
alkoxides and alcohols
+ indicates reaction, - indicates no reaction
2.14 Synthesis of 1-Hz
The carbenes prepared in this thesis were investigated by core
electron spectroscopy in
the group of A. Hitchcock at McMaster University.
Photo electron spectroscopy is a method that allows to determine
the binding energies of
electrons in individual occupied orbitals. Core electron
spectroscopy [45], on the other hand,
allows to map unoccupied orbitals. This method involves the
excitation of core electrons into
the unoccupied bonding or anti-bonding orbitals.
-
68
The energy and the intensity of the transition, for instance,
allow to determine whether a
x* orbital is localized or delocalized.
For the purpose of cornparison of spectml data, the synthesis of
1-H2.was atternpted.
Reaction of [1-H] CI with BH3-THF, LiAIH4 or catalytic
hydrogenation of 1 (H2 / Pd) was
studied. The reaction of [l-HI Cl with BH3 gave a broad product
spectrum and was not
investigated further. With LiAIH4, 1-H2 was obtained but
decomposed whenever attempts were
made to isolate the compound (Fig. 68).
118 LiAlH, THF 'BU I
abH 25 OC 5 min N - LiCI
- [*yH I
N H
'BU I
'BU
(1-HI CI
Fig. 68. Synthesis of 1-Hq
In solution, 1-H2 is stable indefinitely (fiame-seded NMR tube)
but decomposes rapidly
upon exposure to minimal traces of air or moisture. Thus, al1
attempts to isolate the compound
in pure form only led to greyish solids that are presumed to be
polymeric bH2. The instability
of l w H 2 is surprising in view of the fact that a number of
derivatives e. g. [1-a+, 1=0 and l=S
are thermaily robust and can be isolated by sublimation without
difficulties. It is likely that the
N-CH=CH-N fragment is stable only in conjugation with -M
substituents such as C=O, C=S
etc. The sila analog L'SiH2 is also very air-sensitive but can
be isolated by distillation (M.
Denk, personal communication).
-
69
2.15 Reaction of 1 with Fe(C0)s
The stable silylene, L'Si:, readily reacts with metal carbonyls
to give silylene complexes
(L'Si)2Ni(C0)2 [46], L'Si=Fe(CO)q (unpublished results) and
other L'Si-metal complexes.
To establish the relative reactivity of stable silylenes and
stable carbenes, the reaction of
1 with Fe(C0)S was investigated.
With electron-rich carbene ligands, thermal disproportionation
has been found to yield
biscarbene complexes (Fig. 69) [47,48] .
M = Cr, Mo, W
Fig. 69. Synthesis of a bis-carbene complex
The reaction of 1 with Fe(CO)5 (1:2) resulted in the formation
of a yellow, powdery
solid that was isolated by sublimation at 70 O C in vacuo in low
yields and starting material 1
that sublimed at 60 OC in vacuo. The CO stretching frequencies
(V = 2024, 19 12, 1898 cm- l ) of
the yellow sublimate at 70 OC indicate the presence of terminal
carbonyl groups. The
sublimation residue (brown solid) did not show CO bands.
1 1 ~ F e ( c 0 ) ~
Fig. 11. Attempted synthesis of l=Fe(COh
-
70
Based on the unusually high volatility (80 OC I 0.1 Torr), the q
S carbene complex
1=Fe(C0)2 is more likely than the ql complex (Fig. 70). This is
supported by the IR specinim
mat shows two strong bands (v(C0): 1898 (asym.), 1912 (sym.)
cm-').
'BU ?
1 1 =Fe(CO)4
Fig. 70. q l complex formation: l=Fe(CO)4
Attempts to crystallize the yellow sublimate failed both by
layering a 1 :2 solution of
THF and hexanes and by heating it in a flame-sealed NMR tube to
1 10 OC (oven) over a period
of 7 to 14 d. The sublimate transformed into a brown tarry
material upon heating. Due to it's
poor solubility in C&, NMR was measured in THF with a drop
of TMS and a D20 insert as a
lock solvent. The shifts for rert-butyl carbon and the
quaternary carbon were overlapped by
strong THF signais (26.38 ppm, 68.22 ppm). Signals at 129.01 ppm
(N
-
71
After heating at 110 OC for 16 h. dong with signals for 2 and
2-CHO, new signals at
1.17+ 1.5 1 (int. ratio 9 : l), 5.25+6.80+7.74 (int. ratio 2 : 1
: 1) developed. No change was seen
in the spectmm after continuing to heat for 7 d. The appearance
of new signals indicated a
reac tion .
Fig. 71. Aaempted synthesis of 2=Fe(CO)4
When repeated on a larger scale (0.14 g 2 + 0.30 g Fe(C0)5),
different results were different. The FT-IR data of the crude
mixture suggested the presence of an Fe(C0)2 fragment
with two strong CO bands at v = 1848 cm-' and 1908 cm-'.
Sublimation gave two fractions: 45-
50 OC, colorless crystalline, 40 mg and 60-70 OC, purple,
crystalline, 9 mg. Both the volatile
fractions were characterized as 2-CHO by IR specuoscopy. The
sublimation residue (33 mg,
brown powder) showed CO frequencies at 1842 cm-1 and 19 15
cm-'.
2.17 Aromatie Delocalization in Stable Carbenes: Correlation of
Experimental and
Cornputationd data
2.17.1 Introduction
The relative extent of aromatic delocalization in stable
diaminocarbenes (1,3-
imidazolylidenes, 1) and the related species 1=0, l=S and 11-Hl+
was studied at the B3LYP /
6-3 lG* level. The criteria used to evaluate the relative extent
of aromatic delocalization were:
HOMO-LUMO gaps (Eg), bond order (Mulliken and Lowdin), C=C
stretching frequencies and
experimental NMFt data (1H and 1 3 ~ NMR).
-
72
For the derivatives of the carbenes, the C=C stretching
frequencies, the HOMO-LUMO
gap energies and the bond orden lead to different sequences of
increasing delocalization.
corn pou nd l-HZ 110 [I-HI+ 1 6 1 L' B' LW+
V [ C ~ [cm- J 1 706 1 649 164 1 1638 1630 1626 1591
Table 12. Increasing delocalization as obtained fiom
complrtational IR fiequencies
compound 1-Hz 1 4 1 L'B' 1 1 - ~ + LIN+
Bond Order 1.86 1.77 1.72 1 .71 1.68 1 .58
Table 13. Correlation between bond order and NMR data
- - -
corn pou nd L'Ba 1 4 [~-HI+ f L'N+ ~ -HZ
Eg! [eV] 82.00 235.86 248.12 248.2 1 248.87 1591.4
Table 14. Conelation between Eg and aromatic stability
The only criterion that directly correlates with experimental
data is the Lowdin bond
The recent synthesis of stable, diaminocarbenes 1 [6] and the
isostructural stable
silylenes [49], germylenes [SOI and phosphenium cations [5 1,
521 has triggered an ongoing
debate about the extent of aromatic delocalization in these
species. While there is liale doubt
that heterocycles 1 are aromatically stabilized to some extent
or the other, the extent of
delocalization, that is the relative importance of the mesomeric
structure l b us. la , is unclear
(Fig. 4).
R R R I I 1
[*.: R = adamantyl c): - @: mesityl I
metbyl I R R
I R ~ - P ~ O P Y ~
ter& buty l
Fig. 4. Aromatic divalent carbenes la and 1 b and non-ammatic
carbene 2
-
73
The controversy is essentially caused by the need to reconcile
computational data
with experimental data that are obtained from such different
approaches as single crystal X-
ray crystallography, NMR, neutron diffraction and the
measurement of optical and magnetic
properties.
Surprisingly, little attention has k e n paid to the
investigation of delocalization
through computationally derived bond orders in species 1. Bond
orders can only be obtained
by calculations but they can correlate with vibrational data. E.
g. the stretching frequency of
the C=C double bond in heterocycles 1 should correspond to the
bond order and should
decrease with increasing delocalization (Fig. 4).
Examination of experimental IR data for the compounds L'E: (E =
C, Si, Ge) in our
group has shown that the CC-stretching frequencies are weak and
easily obscured by other
strong bands (VCN, VW). The CC-stretching bands are however
easily measured by Raman
spectroscopy and are in fact the strongest bands in the region
2500 - 800 cm-'. Some disadvantages of the experimental approach
remain. Only a srnall number of
possible compounds of type 1 have been characterized and the
precise location of vcc band
can become obscured by Davidov splitting or symmetry enforced
splitting resulting from the
syrnmetry of the solid (space group).
An obvious alternative would be the accurate calculation of vcc.
The calculation of
accurate vibratory spectral data from fint principles is still a
formidable task. Without going
into the details of the problem, the difficulties are arnply
illustrated by an inspection of the
contemporary arsenal of computational methods. Among the
overwhelming number of
different semi-empirical methods: ab initio methods, density
hinctional methods and hybrid
density functional methods, only one approach, the B3LYP hybrid
method, has consistently
delivered accurate vibrational frequencies [8,56].
The accuracy of B3LYP 1 6-31G* calculations is typically better
than 2 W. Even
better agreement between experiment and theory can be achieved
by the introduction of
scaling factors. These factors depend on the type of bond
investigated and can be optimized
by a least squares fit of experimental and computational values
[8].
-
74
In the context of this study, the goal is not to maximize
agreement between
experimental and computational values (vibratory data), but,
rather to study trends in the
delocalization for the family of heterocycles 1 and 2 with
divalent fragments, E.
'Bu I ci-
& N I
'BU
'Bu I
N
I (1-HI Cl 1-Hz 1- l=S
Fig. 72. The stable carbene 1 and its derivatives
B3LYP / 6-31G* calculations are more time consuming than HF /
6-31G*
calculations. Even for R = H, the calculation of the
heterocycles 1 and 2 requires 30 - 40 h of CPU time on a Silicon
Graphics Workstation with a R4400 processor and 128 MB of RAM.
The calculations are useful beyond the goal of obtaining vcc
data. Like other DFi'
methods, the B3LYP method automatically includes electron
correlation. While the effect of
electron correlation on the structural accuracy is not always
clear [ I l ] , electron correlation
can be essential to obtain accurate thermochemical data like
heats of formation etc. The
B3LYP 1 6-31G*