Beta-lactam Synthesis via Gylcol cleavage and Iminization
Tarun Sai Bhadri
Abstract
The breaking of a covalent bond is the heart of every reaction in organic chemistry. In
this experiment the cleavage of 1,2;5,6-Di-O-isopropylidene-D-mannitol was attempted. The
covalent bond between the two hydroxyl groups of the starting material underwent a heterolytic
cleavage, otherwise known as ionic fission. The attempt was then made to convert the main
functional group of this reaction’s product, an aldehyde, into an imine through the nitrogen
source of aniline. This step reflects the contribution of resonance structures to carbonyl
chemistry. This imination facilitated the functionality necessary to form the four-membered ring
that is essential to the beta-lactam-based antibiotic structure.
Introduction
The goal of this experiment was to make a beta-lactam-based antibiotic. Beta-lactam is notorious
for being the structure upon which many antibiotics were made,
with its iconic 4-member ring including nitrogen and a carbonyl.
Beta-lactams work by binding to penicillin-binding proteins (PBP),
the proteins that synthesize the cell wall of many bacteria, and
inhibiting bacterial cell wall synthesis. The most well-known of
which include penicillin, which was first used against
Staphylococcus aureus. The method chosen to synthesize the
beta-lactam was because it used many aspects of carbonyl and
acid-base chemistry in the steps. Alcohol oxidation to aldehyde
and imination reactions are very common when synthesizing
beta-lactam structures because of the relatively easy-to-find reagents.
Experimental 1
1,2;5,6-Di-O-isopropylidene-D-mannitol (5 g, 0.02 mol) was added to a 250 mL
round-bottomed flask with a stir bar and a greased reflux condenser attached. Then
dichloromethane (50 mL) was added to the round-bottomed flask. The reaction was allowed to
reflux until evenly dissolved at 43℃ in an oil bath. The temperature was then decreased to 35℃
and the condenser was removed. Sodium periodate (8.56 g, 0.04 mol) was dissolved in sufficient
amount of deionized water (~100 mL) then added to the round-bottomed flask. The reaction was
allowed to run for an hour and monitored by TLC using methanol. The reaction was then dried
over anhydrous magnesium sulfate (~2.5 g) and stirred for an additional 15 minutes. The mixture
was then filtered using glass wool and LLE was performed, washing with DCM (2 x 10 mL).
The solvent was removed by rotary evaporation at 40℃ and the oil product was characterized by
1
H-NMR and IR spectroscopy.
Scheme 1. Cleavage of Starting Material
Results and Analysis for part 1
Figure 2a. Image and labels of intended product
Figure 2b: Table of 1H-NMR peaks of
Figure 2c
1
H-NMR Hydrogen atom
(ppm)
1.20 - 1.71 HDO
3.76 - 3.98 B
4.02 - 4.28 C
5.22 D
9.66 - 9.72 A
A 1H-NMR spectrum (Figure 2c) was done in order to determine the presence of an
aldehyde. The peaks were identified and matched to the hydrogens of the expected product
(Figure 2a) and are organized in Figure 2b. The large peak at 1.20 - 1.71 ppm is due to water as a
result of inadequate drying in the purification process. Hydrogens B are not as affected by the
electronegativity of the two oxygens so they are more shielded than C and D. The two C
hydrogens are near an oxygen atom so they will be more deshielded because of the oxygen’s
high electronegativity. The D hydrogen atom is near an oxygen atom and an aldehyde causing it
to be more deshielded than C because of both of these oxygen withdrawing effects. Finally,
hydrogen A is an aldehyde so it is extremely electron deficient and characteristically shows up
around 9.7 ppm.
An IR spectrum (Figure 2d), was taken to analyze the functional groups in the product.
The peaks at 3422.52, 2986.59, and 2936.11 cm-1 are likely due to water in the product that was
not removed by the MgSO4. The peak at 1735.16 cm-1 further correlates the presence of an
aldehyde in the isolated product. However, due to the water signals appearing in the 1H-NMR
and IR spectra, it is clear that the product is not pure.
Discussion 1
The percent yield for experiment one was 4%. This was most likely due to the fact that
the aldehyde structure is very reactive and could dimerize back to the starting material relatively
quickly when it is not stored at -78℃. Unfortunately, the product was stored in a freezer at 0℃.
Additionally, the purification method, using silica gel, may have reduced with the yield, as the
hydroxyl groups on the silica gel may have reacted with the aldehyde,
Experimental 2
1,2;5,6-Di-O-isopropylidene-D-mannitol (3g, 0.011 mol) were added to a round bottom
flask with a stir bar and condenser. Next, 30 mL of ethanol was added and the mixture was
heated at 83°C under reflux until completely dissolved. The temperature was decreased to 35°C
and the condenser was removed. In a round bottom flask, sodium periodate (5.13g, 0.024 mol)
was dissolved in 50 mL of deionized water. This solution was then added to the initial round
bottom flask ands stirred for one hour. Next, the reaction mixture was filtered and washed with
two 10 mL washes of ethanol. Then, magnesium sulfate was added and the mixture was filtered
again. The solvent was removed by rotary evaporation and the resulting product was analyzed
through IR spectroscopy and 1H-NMR. DCM was substituted with ethanol because it is miscible
with water and this was needed to ensure that the sodium periodate remains dissolved in the
reaction mixture.
Scheme 2. Cleavage of Starting Material Using Ethanol
Reaction Mechanism 2
Results and Analysis for part 2
Figure 3a. Intended product for experiment 2
Figure 3b: Table of 1H-NMR peaks of Figure 3c
1
H-N Hydroge Explanation
MR n atom
(ppm)
1.07 A These hydrogens are part of a methyl group with an expected shift of 0.9. They
are shifted downfield do to their proximity to two very electronegative oxygen’s
which deshield the hydrogens in the methyl groups.
1.19 B These methyl groups have an expected shift of 0.9. They are shifted downfield
do to their proximity to two very electronegative oxygen’s.
2.23 E These two hydrogens have an expected shift of 1.3 but are shifted downfield
because of their proximity to two oxygens.
3.57 C This methyl group has an expected shift of 3.8 because it’s placement is similar
to an ester. It is shifted upfield because it is shielded from the oxygen by a CH2
group.
3.67 D These two hydrogens have an expected shift of 3.8 because they resemble an
ester. They are shifted upfield due to shielding produced by the adjacent methyl
group.
4.79 HDO This large peak is due to water as result of inadequate drying.
The hydrogens labeled E are diastereotopic. The hydrogens labeled A, B, and C are
homotopic. The hydrogens labeled D are enantiotopic. A 1H-NMR spectrum (Figure 3a) was
done in order to determine the presence of an aldehyde. The most important H-NMR shifts to be
expected was in the 2-2.3 ppm range and would have indicated the presence of the aldehyde.
Shifts near 0.9 ppm would have indicated the methyl groups and a shift near 1.3 would
correspond to the hydrogens within the ring.
An IR spectrum (Figure 3d), was taken to analyze the functional groups in the product.
The most significant IR peak was expected in the 1720-1730 cm-1 range and would have
supported the presence of an aldehyde. The experimental peaks did not support the intended
product. The peak at 3347 cm-1 is broad and strong so it provides evidence for the presence of a
hydroxyl group. The peak at 1041.26 cm-1 is sharp and medium which is evidence of an ester.
Discussion 2
The H-NMR and IR data support the conclusion that a hemiacetal was formed instead of
an aldehyde. This may be because the substitution of DCM with ethanol provided a nucleophile,
ethanol, which was able to attack the carbonyl. This means that resonance would have allowed
the carbonyl to become an electrophile within water. This hemiacetalization explains the absence
of a carbonyl peak from the IR spectroscopy.
In the subsequent attempt of this reaction, ethanol was substituted with DCM to eliminate
the presence of nucleophiles and preserve the aldehyde needed for the imination reaction.
Experimental 3
1,2;5,6-Di-O-isopropylidene -D-mannitol (5 g, 0.020 mol) was added to 50 ml of
anhydrous dichloromethane (DCM/CH2Cl2) in a 250 mL round-bottom flask with a stir bar and a
condenser. The flask was heated to reflux until the starting material was evenly dissolved in the
solvent. Then the flask was cooled via ice water bath, and the condenser was removed. While
stirring the flask, saturated aqueous sodium bicarbonate was added, (0.5 mL/g, 2 mL) and then
sodium metaperiodate (NaIO4; 8.56g, 0.040 mol, 2 equiv.) was added. Flask was kept at below
35 degrees Celsius with continuous stirring for one hour. The reaction was monitored by TLC
and 1H-NMR in anhydrous CDCl3. The reaction mixture was gently filtered through a pad of
Celite to remove NaIO4 and the sodium bicarbonate. A final 1H-NMR and IR were taken to
confirm the presence of the aldehyde.
Then in the same flask with the product from step 1 (0.036 mmol), MgSO4 (0.074
mmol, 2.0 equiv.) was added. Then, the flask was stirred for 3-5 mins. Anhydrous aniline (0.10
mmol or excess) was added to the round bottom flask. The reaction was allowed to stir for 30
minutes and monitored with TLC. A final 1H-NMR and IR was taken to confirm the presence of
the imine.
Scheme 3. Aldehyde formation using sodium bicarbonate
Scheme 4. Imination
Results and Analysis for part 3
Figure 4a. Expected product after 30 min
Figure 4b: Table of 1H-NMR peaks
of Figure 4c
1
H-NMR (ppm) Hydrogen atom
1.33, 1.35 B
4.23, 4.03 C
5.24 D
1.40 HDO
9.63 A
1
H-NMR (Figure 4c) was performed 30 minutes into the first step of this experiment in
order to determine if an aldehyde peak was present. An IR spectrum was not performed at this
time because it would require too much of the product which would dramatically reduce the
percent yield. The peaks at 1.33 and 1.35 ppm indicate the methyl groups (Hydrogen B) attached
to an sp4 carbon. The diastereotopic hydrogens (C) are present in the esteric shift of 3.8 ppm and
are shifted upfield due to shielding produced by the adjacent methyl group. Hydrogen D was a
very strong peak at 5.24 ppm and is less shielded due to the neighboring oxygen. 9.63 ppm most
likely resembles aldehyde hydrogen (A).
Figure 5a. Expected product after 1 hour
Figure 5b: Table of 1H-NMR peaks
of Figure 5c
1
H-NMR (ppm) Hydrogen atom
1.33, 1.35 B
4.23, 4.03 C
5.24 D
1.40 HDO
9.63 A
Figure 5e. The Starting Material
Once the experiment was allowed to run for an hour, an IR spectrum (Figure 5d) and
1
H-NMR (Figure 5c) were taken in order to determine if the expected product (Figure 5a) was
isolated. The IR (Figure 5d) did not have a peak between 3200 and 3500 cm-1 and it does have a
single sharp peak at 1734.88 cm-1 which provides strong evidence that an aldehyde, most likely
the desired product (Figure 5a) was formed. The starting material (Figure 5e) has two alcohol
groups so if this reaction was unsuccessful, IR peaks would be expected around 3200 and 3500
cm-1. The 1H-NMR spectroscopy looks very similar between the 30 mins into step 1 and the end
of step 1 itself, the aldehyde peak is prominent in the 1H-NMR and IR. The peaks at 1.33 and
1.35 ppm indicate the methyl groups (Hydrogen B) attached to an sp4 carbon. The diastereotopic
hydrogens (C) are present in the esteric shift of 3.8 ppm and are shifted upfield due to shielding
produced by the adjacent methyl group. Hydrogen D was a very strong peak at 5.24 ppm and is
less shielded due to the neighboring oxygen. 9.63 ppm most likely resembles aldehyde hydrogen
(A).
Figure 6a. The Expected Imine from Step 2
Many of the 1H-NMR peaks expected for the imine product are not found in Figure 6c.
There was expected to be a strong peak between 7 and 8 ppm for the aromatic hydrogens (A).
However, the peak for I is present at 6.72 similar to a deshielded aldehyde. The peaks for
diastereotopic hydrogens C and D are, respectively, 4.00 and 4.12, and hydrogens E and F show
up at 1.35 as the shifted down methyl groups, due to the single-bonded oxygens. The lack of the
aromatic or benzylic peaks indicates that the imination reaction did not occur, and the desired
product was not made, and the aldehyde product is still present.
Discussion 3
For experiment three, great precaution was taken to ensure that the aldehyde would not
react with anything else in solution and maximize yield. The most important steps that were
taken, was ensuring that there were no nucleophiles or acids in solution to prevent any additions
to the volatile aldehyde product. As a result of this exclusion, anhydrous DCM, aniline, and
CDCl3 were used to limit the amount of water in the reaction to prevent that from adding to the
aldehyde. Since sodium periodate was necessary for the reaction to proceed and is only miscible
in water, saturated aqueous sodium bicarbonate was used in place of brine in the previous
procedures. Sodium bicarbonate is a better replacement because it is a weak base so it will
neutralize any acids in the solution and prevents excess water in the solution. Magnesium sulfate
was removed from the procedure for step one because it would pull all the water from the
solution which would prevent the sodium periodate from reacting with the starting material
which was in the organic layer.
The previous procedures were also edited to improve the purification process since that
likely was a source of error. The sodium periodate and sodium bicarbonate were removed by
filtration. A celite pad was used for filtration because silica is slightly more acidic and polar, so
celite was used to err on the side of caution. Rotary evaporation was used previously to remove
any solvent leftover in the solvent; however, the product has a relatively low boiling point so
there would have been a risk of losing the product. Instead of this, nitrogen gas was blown over
the product to remove the solvent. Nitrogen is nucleophilic, but nitrogen gas is very stable to the
the triple bond holding it together so it should not act as a nucleophile.
Future directions and additional data
Since the first step was one of the most difficult due to the volatility of the aldehyde
produced, increasing the yield and still producing the aldehyde proved challenging. In the future,
mass spectroscopy would have been useful since it was difficult to analyze the data at times to
determine if the product isolated was the expected product. Since the first step involved cleaving
the starting material, a mass spectrometer would have been beneficial for determining the
chemical formula of the compound.
In addition to this, a longer lab period would have also benefited the first reaction. Since
the aldehyde can easily reform the starting material, the second step has to be performed almost
immediately in order to ensure that the amine can react with the aldehyde to form an imine. Due
to this, the second reaction was only allowed to run for 30 minutes. While this should have been
enough time for the amine to react, more time would likely improve the yield of the reaction.
In terms of error, there were a few mistakes made through this process. For the first
experiment, the product of the first step, the aldehyde, was left in a lab drawer at room
temperature for a week. In addition to this, magnesium sulfate was used with sodium periodate,
this prevented the sodium periodate from mixing with the organic layer so the mixture was not
allowed to fully react. Since there was not a lot of product obtained in this experiment, there was
only enough product to run an 1H-NMR (Figure 2c) and IR (Figure 2d). In the second experiment
run, the goal was to assist in the mixing of the sodium periodate and the organic layer. However,
ethanol was used as the solvent in place of DCM which resulted in an addition to the carbonyl
rather than isolating an aldehyde. Finally, the second step of experiment three, transforming the
aldehyde into an imine, the values in the procedure were in mmol rather than mol so there was
not enough product isolated.
References
Kong, K.; Schneper, L.; Mathee, K. APMIS 2010, 118, 1-36, PubMed Central PMID=20041868.
[Link] (accessed April 21, 2019).
Kamath, A.; Ojima, I. Tetrahedron 2012, 68, PubMed Central PMID=23264702
[Link] (accessed April 21, 2019).