123713AB lecture01

These are the old slides that made up the
‘traditional’ version of these two units
(asymmetric synthesis & total synthesis).
I will annotate these slides and see if they
work as the reading material for the course ...
bear with me, it is a bit of an experiment.
These notes are NOT comprehensive but
supplement your own reading. It is
impossible to cover these two areas in just 10
lectures (the original length of this module).
Some of my colleagues would go as far as
saying “we don’t”. They would, of course,
be wrong. There are two quick answers:
1) We need organic compounds so we need
to learn how to make organic molecules.
2) Research and Education. The problems
encountered in total synthesis push
forward the development of new
methodology and teach us the application
of chemistry.
1

It is made by Roche and at least one of
their published routes is based on the
conversion of shikimic acid (isolated from
star anise) to the final drug.
The entire pharmaceutical industry and
much of the agrichemical industry (and
many other industries) is build on the
chemists’ ability to synthesize molecules
with specific properties.
On this slide we see the molecule
oseltamivir or Tamiflu, an antiviral
medication used in the prevention and
treatment of flu.
2

The reported route takes 12 steps to
convert shikimic acid to the final product.
There are many other reported syntheses in
the literature. Some are longer, some are
shorter. We need to learn how to compare
these routes, to determine the pros and
cons of each route and, ultimately, how to
design similar syntheses so that we can
make our own target molecules.
Atorvastatin or Lipitor was one of the most
successful of the statin drugs used to low
blood cholesterol. From 1996-2012 it was
the world’s best selling pharmaceutical. It
is now off patent.
Yet again, it was initially discovered by
organic chemists ...
3

The original synthesis started from
isoascorbic acid, a stereoisomer of vitamin
C. It is a cheap ‘chiral pool’ natural
product.
Again, we need to be able to deduce how
Lipitor can be prepared from this
material ...
I could go on and on ... there are many
examples of small organic molecules used
to treat ailments. This is another bestseller
from the pharmaceutical industry and is
used to both treat and manage asthma.
The synthesis of the top molecule
salmeterol is relatively quick, taking just ...
4

... six steps from this salicylic acid
derivative.
Of course, not all treatments are small
organic molecules. Peptides are becoming
popular targets. Fuzeon or enfuvirtide is a
biomimetic peptide that confuses the HIV
virus. It is hard to make and a tad
expensive ($25,000 US per year) so is a
last resort medicine.
5

But chemists are everywhere. To display my
Massey card, the next example is from both
the agrichemical sector and companion
animal health sector. Imidacloprid is
probably the world’s most commonly
employed insecticide.
On the downside, it has been linked to
colony collapse disorder in bees.
Its synthesis is shockingly simple and it
can be prepared in just four steps from 3-
methylpyridine.
6

Another example from the companion
animal health sector is the various
components of Drontal ...
... while we use this to keep our pets in
good health, one of the major components,
praziquantel, is found on the WHO Model
List of Essential Medicines needed for
basic health care. It used to treat intestinal
parasites.
New synthetic methodology is needed to
allow a cheaper synthesis to be developed
so there is more access to such useful
drugs.
7

Given enough time and resources (money
and students!) it looks like most molecules
can be prepared. One of the most
ambitious syntheses was the preparation of
palytoxin. This is the largest (non-
polymeric, non-peptide) compound I could
find.
It is not entirely clear how many steps are
involved as the synthesis of the starting
materials was not reported.
Interestingly, it is often small molecules
that are hard to prepare. The problem with
such molecules is they lack functionality for
us chemists to play with.
This is octanitrocubane. It was predicted to
be one of the most potent carbon based
explosives but it turns out it is not and that
the heptanitrocubane is more explosive.
This fact shows the importance of shape
and conformation to reactivity (as we shall
see later).
8

A knowledge of organic chemistry is also
required for disciplines such as
nanochemistry; making carbon-based
materials underpins this area. Here we see
a single aromatic belt or a single strip of a
nanotube.
This was made in a rational manner (unlike
most nanotubes which are effectively
purified soot).
So why do we need asymmetric chemistry/
synthesis?
Hopefully you are all aware that the shape
of a molecule is key to many of its
properties. One of the most famous
examples is that of thalidomide (shown
above). This was sold as a mixture of
enantiomers (hands) to treat morning
sickness (amongst other ailments) but
caused children to be born with
malformation of the limbs.
9

There are many examples of different
enantiomers (hands/mirror images) of
molecules having different biological
properties.
This example shows a cough-suppressant
and a painkiller. As you can see they are
mirror images (as are their trade names …)
Stereochemistry also finds its way into
nanochemistry. Above is an example of a
molecular motor that can only rotate in one
direction (and we can control the speed of
rotation by varying temperature and light
sources).
This remarkable molecule has multiple
stereogenic elements including stereogenic
centres (chiral tetrahedral carbons) and a
chiral helix.
10

A similar motor has been used to rotate a
liquid crystal and thus cause …
… a glass rod to be rotated in a specific
direction.
Obviously I can’t put the movie into a pdf
file but I recommend you look it up (or I
might attempt to upload it to Stream but
don’t hold your breath).
11

I keep all of my lecture notes and course
material on my own website. Often you will
find different versions of the current
material. I rarely remove older versions as
some times my attempts at improving
material is not successful (I’m an
experimentalist after all).
My notes are not comprehensive but for the
background to your studies. I recommend
that you consult …
… textbooks
I can’t recommend these books enough for
those of you that have an interest in
organic chemistry, total synthesis and
stereochemistry. They are all very well
written and all very useful.
… and of course …
12

… the original literature.
For the most part I have given the
references to the original literature
(especially for the examples of and from
total syntheses).
Reading the literature is invaluable. All the
major organic journals (some shown above)
will have very good examples of both Total
Synthesis and asymmetric synthesis
(especially catalysis).
We will start with Asymmetric Synthesis.
The rest of this pdf file will look at
introducing some of the basic terminology
that surrounds (and mystifies) asymmetric
synthesis before looking at substrate
control.
I’m going to make the assumption that you
have studied some basic organic chemistry
and that you have encountered
stereochemistry and chirality before …
13

Chirality is not a chemical principle or an
exclusively chemical property. Any object
that cannot be superposed upon its mirror
image is called chiral.
Classic examples are shoes and, of course,
our hands (chiral comes from the Greek for
hand ‘kheir’ and hence pronounce it with a
‘k’ not ‘ch’).
Note: it is superpose not superimpose (which means lay one thing over
another so that both are still evident)
There are many examples such as
corkscrews, propellers and golf clubs …
14

Molecules can be chiral. If the mirror
image of a molecule cannot be superposed
upon the original molecule however many
times you twist or rotate it (and its bonds).
Common examples of chiral molecules are
the amino acids (with the exception of
glycine). The example above is cysteine.
The two non-superposable stereoisomers
are known as a pair of enantiomers or an
enantiomeric pair. A molecule can only ever
have one enantiomer (its mirror image or
twin).
Stereoisomers are isomeric molecules that
have the same constitution (same atoms)
and have the same sequence of bonds but
differ by the three-dimensional
arrangement of the atoms. The most
obvious examples are alkenes.
15

If you have a 1:1 mixture of enantiomers it
is termed a racemic mixture or you could
be said to have a racemate.
Converting one enantiomer into its mirror
image (a process that can only occur by
breaking and making bonds) is called
racemisation.
The opposite of a chiral object or molecule
is …
… an achiral object.
The mirror image of a spoon can be
superposed upon the original object and so
it is an achiral object. The two mirror
images are identical (there is no twin).
16

This is the same for molecules. If the
mirror image of a molecule can be
superposed upon the original molecule it is
achiral and has no ‘twin’. The two mirror
images are the same molecule.
The simplest test for an achiral molecules
to see if it has a plane of symmetry
(internal mirror plane). Imagine a mirror
dissecting the molecule in half, if both
sides are identical (one side is reflected on
the other) the molecule is achiral. If it does
not have a plane of symmetry it is chiral.
This is simplistic but works for the vast
majority of molecules.
I think the formal definition is that a molecule is achiral if it has an improper
rotation, which might take into account other forms of symmetry.
17

We need to be able to communicate the
chirality of a molecule with each other.
There are many (historical) ways of doing
this. The organic naming convention now
uses the Cahn-Ingold-Prelogue (CIP)
system for stereocentres (and it is often
used for axial, planar & helical chirality,
even though helical chirality has a different
system, which can be employed for axial
and planar chirality … it’s confusing).
Inorganic chemists use a different system.
This should be revision so I’m not going
through it. If you don’t know how to assign
R or S look at lecture notes for 123.202.
Basically, to designate a sign to a chiral
centre look at its substituents. The higher
the atomic number the higher the priority.
Keep moving along the substituent until
you find a difference.
This done for you with alanine (above).
18

Once you have assigned the four priorities
place the lowest priority (number 4 (4th) by
convention) away from you.
This may involve rotating the molecule. Be
careful, this is where most mistakes are
made.
The draw an arrow connecting the
remaining three substituents, going from
highest priority to lowest priority.
If the arrow is clockwise (or to the right)
the stereocentre is designated the
descriptor R. If it is anticlockwise then it
has the descriptor S.
So if we return to alanine …
19

… you should see that the arrow is going
anticlockwise so this amino acid is given
the descriptor S.
Please remember that this designator is
not based on any physical property of the
molecule (rotation of plane polarised light
etc.) and is purely a very useful part of
chemical nomenclature.
The CIP rules work for all stereocentres and
not just those based on carbon.
So we can use R or S for the sulfoxide (if
you do not understand why a sulfoxide is
chiral then please draw the Lewis dot &
cross structure) and phosphine oxide above
(both are S if you interested).
The CIP rules can be applied to axial and
planar stereoelements but rarely helical
chirality (or inorganics).
20

A molecule is either chiral or it is not. It can
either be superposed upon its mirror image or
not. But chemists don’t like things so simple. So
we class these molecules depending on the motif
that breaks the symmetry of the molecule.
The example above is a stereocentre (but often
termed an axis of chirality. Please note that the
stereocentre is a tetrahedral carbon with only 2
different groups so ignore what you were taught
as an undergraduate … again.
Called oleane, the molecule above is the sex pheromone of the
olive fly.
The classic example of axial chirality is
BINAP.
This does not contain any tetrahedral
carbon atoms but is still chiral. Rotation
around the bialy C–C bond is restricted (not
free to rotate). This results in two non-
superposable mirror images (like scissors).
Restricted rotation around a single bond as
atropisomerism. There is a good review by Clayden in Angew.
Chem. Int. Ed. 2009, 48, 6398-6401.
21

Helical chirality is beautiful to look at. Often
described as inherent chirality as the
molecule is chiral regardless of any
substituents (most chiral molecules are only
chiral as a result of substituents attached to a
central point, inherent molecules are curved).
Interestingly, helicenes often undergo
racemisation at a lower temperature than one
might imagine due to a cycloaddition-
retrocyclisation process.
My personal favourite form of chirality is
planar chirality as exhibited by ferrocene
derivatives and, of course,
[2.2]paracyclophane derivatives.
22

A molecule with a single stereogenic
element can only exist as two
stereoisomers, the enantiomers.
If a molecule has two stereogenic elements
then it can have up to four stereoisomers. It
will have one mirror image. The mirror
image has the same relative
stereochemistry (both substituents on the
same face) but different absolute
stereochemistry (are they up or down).
It will also two diastereoisomers. These are
molecules that have a different spatial
arrangement of their atoms
(stereoisomers) but they are not mirror
images.
They have the different relative
stereochemistry (the relationship between
substituents) and different absolute
stereochemistry.
23

There are another pair of enantiomers
(mirror images or twins).
Enantiomers as mirror images have almost
identical properties (as you would expect).
Diastereoisomers are completely different
molecules so have completely different
properties.
So if a molecule has two stereogenic
elements it can have up to four
stereoisomers.
24

Here we have the stereoisomers of the
aldopentoses. These molecules have 3
stereocentres. There are 8 possible
stereoisomers. Each molecule has 1
enantiomer and 6 diastereoisomers.
Another way of describing this would be
four enantiomeric pairs.
It turns out that the maximum number of
stereoisomers a molecule can have is given
by the formula 2n, where n = the number of
stereogenic elements.
But …
25

… we have to remember that this is just the
maximum possible number of
stereoisomers. There might be less …
Here is tartaric acid (or, to give it its old
name, racemic acid. And yes, it is
important in the history of asymmetric
synthesis).
At first glance it might appear that there is
four stereoisomers on this page but there
isn’t …
26

Hopefully, you can see that all of these are
diastereomers of each other (both the
relative and absolute stereochemistry are
different for each molecule).
Hopefully, you can see that these two are
enantiomers. They are mirror images of
each other and they cannot be
superimposed.
It is the other two that we have to inspect
in more detail …
27

… these two are actually identical.
They may look like enantiomers with each
molecule containing two stereoisomers …
… but if we inspect the molecules in more
detail we find that they contain a plane of
symmetry (which is shown on the next
slide). This means they are achiral so will
be superposable upon their mirror image.
Or, in other words, they are identical.
Molecules that contain multiple stereogenic
elements but are still achiral are called
mess molecules.
28

The internal mirror plane means the
molecule is achiral.
Note: it is really important to be able to
manipulate our representations of
stereocentres.
Remember that an sp3 carbon is
tetrahedral. Only two bonds can be in the
plane of the paper at one time.
To change which bonds are in plane (but
not swapping any groups around) simply
remember that if you start with a bond up
you will end with a bond up.
29

Alternatively, if you fully rotate the C–C
bond shown this cause the two groups at
the end of the bond to swap places. If you
do this then you must also swap a wedge
for a dash (or vice-versa).
Please note these are ‘cheats’ or
stereochemical hacks; try to picture the
molecules in your head (or draw them) and
understand the changes during any
manipulation.
Enantiomers will have identical properties.
They are mirror images so they are
energetically identical. As a result all
physical properties such as mp, ir and nmr
will be identical for a pure sample of each
molecule (and yes that word ‘pure’ is very
important).
30

Diastereoisomers are not related by
symmetry. They are not mirror images.
They are totally different molecules so they
can have completely different properties.
You would not expect structural isomers to
have the same properties so you should not
expect diastereomers to have the same
properties.
Diastereomers are different molecules.
The only difference between enantiomers is
their interaction with other chiral objects as
the interaction forms a diastereomeric
adduct/intermediate/complex whatever.
The difference between diastereomers is
key to asymmetric synthesis.
We have form diastereomeric interactions
or transition states if we are to measure,
separate or synthesise stereoisomers.
31

So how do we measure the purity of an
enantiomer if its physical properties are
the same as its mirror image?
Well before we answer the how we measure
the purity just a quick mention how we
report the purity …
The most common measure of enantiomer
purity is ee or enantiomeric excess. It is an
archaic measurement connected to optical
purity (measuring the rotation of plane
polarised light).
Enantiomeric excess is a measurement of
how much more of one enantiomer than the
other we have. Basically, its says the minor
isomer cancels out 20% of the major isomer
(80%) and we measure what is left (60%).
32

The same two reporting methods are used
to indicate diastereomeric purity.
Diastereomeric excess seems really
pointless when talking about the formation
of diastereomers as all measurements
actually give the ratio. Unfortunately, you
will still see it in the original literature.
A more modern variant (that should be
used more often) is enantiomeric ratio er.
The version of er normalised to a value out
of 100 (e.g. 80:20) can be used in
calculations more readily than ee.
Furthermore, the methods of measurement
effectively give this value directly.
33

So how do we measure enantiomeric ratio
or enantiomeric excess?
Measuring the purity of an enantiomer is
hard. Remember virtually all physical
properties of each enantiomer is identical.
It is like having a mixture of left & right
handed gloves. They are identical in every
way; they made of the same material, have
the same number of fingers and thumbs.
You cannot tell the difference between the
two … unless …
34

… you add a chiral object such as your left
hand.
Now it is clearly obvious if the glove is left
or right handed. Only one will fit.
The reason the gloves are now different is
that there is a diastereomeric interaction
between two chiral objects (glove and
hand).
Molecules are just the same. The only way
we can measure how much of each
enantiomer we have is by studying the
interaction of our sample with something
chiral thus creating diastereomeric
interactions.
The second source of chirality could be a
molecule, a machine or special forms of
light.
35

In cartoon form; we take our mixture of
enantiomers (the circles). They are identical so
can’t be separated or measured until they
interact with an additional source of chirality
(the square). This creates diastereomers (circle
+square). Diastereomers are different so we can
measure or separate these.
The separation of enantiomers is often called
resolution.
Over the next slides we will look at different
methods of achieving this …
The first method is derivatisation. This is where
we form a covalent bond between our substrate
and an enantiomerically pure molecule.
The classic example is the addition of the
Mosher acid (to make an ester or amide).
The new diastereomers are often separable.
They can normally be observed in the 1H & 19F
NMR and they can be used to determine the
absolute stereochemistry of secondary alcohols
of unknown configuration. A possible
disadvantage is the addition of extra steps.
36

You do not have to form a covalent bond. If
the substrate contains an acidic or basic
functional group then diastereomeric salt
formation can be useful. Resolution is then
achieved by selective crystallisation.
This is commonly used in the preparation
of pharmaceuticals (as shown above) and
many common chiral ligands.
Again it adds additional steps …
It is simpler not to derivatise the substrate
but rely on temporary interactions (H-
bonding, pi-stacking, van der Waals forces
etc) to resolve the enantiomers.
HPLC or GC with a chiral column (stationary
phase) is the most common method to
measure enantiomeric enrichment.
The problem is cost, finding the correct
conditions/column and that it is rarely used
on a preparative scale.
37

Having recapped the basics lets start
looking at the selective synthesis of pure
enantiomers and diastereomers or the
process normally called asymmetric
synthesis (although this is a slight
misnomer as a molecule can be symmetric
but chiral).
We will start by looking at substrate
control; the use of existing stereochemistry
within a molecule to control the
introduction of a new stereocentre.
The first example is taken from a synthesis
of canadensolide … a natural product with
anti-fungal properties …
38

This is an example of a Mukaiyama aldol
reaction (the Lewis acid-mediated addition
of a silyl ketene acetal to an aldehyde).
In this reaction an achiral nucleophile is
attacking a chiral electrophile. (if you do
not know the curly arrow mechanism
please start revising your organic
chemistry).
We can see the addition gives exclusively
one diastereomer.
The question is “why is only one
diastereomer formed?”
39

This should be revision from
undergraduate.
There are a number of models that predict
the substrate controlled facial selectivity of
the addition of a nucleophile to a carbonyl
group. I will not describe all of them.
One of the most useful is the Cram
Chelation Control model …
It is assumed that a nucleophile will
approach a carbonyl group along the so
called Bürgi-Dunitz angle (~107°). This
maximises orbital overlap of the HOMO of
the nucleophile and the LUMO (𝛑*) of the
carbonyl (while avoiding repulsion from the
C=O HOMO). (It also relates to the transition state and the fact the
carbon will soon be sp3)
This means there is an interaction with the
R groups (ignoring the Flippin-Lodge
angle).
40

As the nucleophile approaches from an
obtuse angle it passes over the carbonyl
substituents and interacts with them.
We need to determine the conformation of
the substrate. In the Cram Chelation
control model the conformation is locked
by a Lewis acid coordinating to the
carbonyl and a Lewis basic substituent
(L=large, S=small & Z=Lewis base). This
prevents free rotation.
The nucleophile then approaches least
sterically demanding face of the carbonyl
(it moves passed the smallest substituent).
This is shown above as either the skeletal
representation or the Newman projection
(please practice swapping between the two
representations).
This indicates which face of the carbonyl
will be preferentially attacked (but not the
degree of this preference).
41

Lets look at a second example of substrate
control.
This is taken from a synthesis of
preswinholide (swinholide is a dimer of this
molecule).
Massey might not have online access to
this journal pre-1996 (but a print version is
in the library … that wonderful building full
of books where you all check your
FaceBook updates … )
Here we observe the Lewis acid-mediated
addition of a weak nucleophile (allyl silane)
to an aldehyde.
Effectively this is the same addition as
before except we have exchanged a
methylene group (CH2) for a oxygen atom.
Again we observe good diastereoselectivity.
42

Again the question is why?
And, funnily enough, it is not going to be as
a result of Cram Chelation control. In this
example the bulky silyl ether protecting
group on the alcohol prevents the titanium
from tying the groups together and fixing
the conformation of the substrate. (Funny
because I wouldn’t ask the question if the
answer was obvious … I’m such a hilarious
character).
Key to predicting the outcome is the
conformation of the substrate. This is given
by the Felkin-Anh model, which is believed
to represent the transition state of the
addition.
The conformation of the substrate has the
largest substituent perpendicular to the
carbonyl group. The nucleophile
approaches along the Bürgi-Dunitz angle
and passes the smallest substituent as
shown above.
43

This model is very useful at teaching us the
basic principles that here is another
example.
How would we predict which of the two
diastereomers was favoured?
Note: We cannot predict how selective the
reaction will be, just which face of the
carbonyl will be preferentially attacked (to
improve selectivity make nucleophile
bigger)
We need to predict the reactive
conformation, the conformation the
substrate adopts during the reaction.
Felkin-Anh says that this will have the
largest group perpendicular to the carbonyl
group (minimise non-bonding interactions).
First we convert the skeletal representation
into the Newman projection. In this case
the phenyl group is in the same plane as
the carbonyl. Then …
44

… we rotate the back carbon atom so that
the phenyl group is perpendicular with the
carbonyl group.
There are two conformations that fulfil this
criterion.
We are only interested in the one that has
the smallest substituent close to the Bürgi-
Dunitz angle (close to the substituent of
the aldehyde/ketone) …
… the nucleophile will attack the face of the
carbonyl opposite large phenyl group and will
approach along the Bürgi-Dunitz angle (107°)
passed the smallest substituent.
The hardest part is converting the Newman
project into a skeletal representation (many
people no longer teach the Newman projection).
I leave the original stereocentre unchanged and
then rotate the front atom so that the longest
carbon chain is anti-coplanar (same plane but
opposite sides). Then look at the relative
stereochemistry.
45

The Felkin-Anh model also explains the
observed results from molecules that
contain a heteroatomic substituent on the
𝛂-carbon.
In this example we see that both
diastereomers can be accessed depending
on the reagent used.
If we use a Lewis acidic reagent that can
coordinate two Lewis bases together …
… then we have an example of Cram
Chelation control and we obtain the anti-
Felkin-Anh product.
If we use a non-coordinating reducing
reagent then we get the opposite result.
At first glance it may not be obvious which
group is the bigger, the isopropyl group or
the methylsulfide but it turns out it does
not matter …
46

The Felkin-Anh model also states that if
there is an electronegative element on the
𝛂-carbon it will adopt the perpendicular
position. This maximises orbital overlap
between the carbonyl group and the C–S
bond. This in turn reduces the energy of
the LUMO and favours nucleophilic attack.
Thus the diastereoselectivity is a balance of
steric and electronic factors.
(the old organic catch-all that generally
means ‘we’re not sure what is going on’ but
for once is actually true)
47

So hopefully this has acted as an
introduction to substrate control. The idea
that an existing stereocentre can control
the diastereoselectivity of a reaction.
Basically, the control comes about because
the two possible transition states for the
addition of a nucleophile to a carbonyl
group are diastereomeric and they have
different activation energies. The transition
state with the lowest activation energy will
be favoured. This invariably means the
nucleophile approaches closest to the
smallest substituent on an adjacent
stereocentre …
Above is a more complex diagram trying to
represent this idea (taken from the next
version of this course, which unfortunately
is not ready for public consumption yet).
Next lecture will look at other substrate
controlled reactions.
48

123713AB lecture01

More Related Content

What's hot (20)

Viewers also liked (12)

Similar to 123713AB lecture01 (16)

More from Gareth Rowlands (9)

Recently uploaded (20)

123713AB lecture01