Chapter 3: Determining Organic Structures

Welcome to Last Minute Lecture.

This free chapter overview is designed to help students review and understand key concepts.

These summaries supplement not replaced the original textbook and may not be redistributed or resold.

For complete coverage, always consult the official text.

Welcome to the Deep Dive.

We're here to cut through the noise and get straight to the crucial insights.

Today, we're tackling a really fundamental question in chemistry.

How do we actually know what a molecule looks like?

It's shape.

It's composition.

I mean, think about it.

You isolate something new, maybe from a natural source, could be a potential medicine.

How do you figure out precisely what it is and fast?

So that's really what we're getting into today.

We're diving deep into determining organic structures, pulling insights from key resources like Clayton, Greaves, and Warren.

It's all about the power of spectroscopy, the essential toolkit chemists rely on every single day.

Yeah, it's difficult to really convey just how much spectroscopy revolutionized organic chemistry.

Before these methods were common, I mean, figuring out a structure could take ages, years sometimes.

Lots of chemical tests, degradation, quite a bit of, well, inspired guesswork, frankly.

Spectroscopy changed all that, speed,

certainty.

We'll be focusing on the big three today.

Mass spectrometry that tells us the mass, the atomic makeup, then nuclear magnetic resonance, which is amazing for mapping out the connections, molecular skeleton.

And finally, infrared spectroscopy, or IR.

That one's brilliant for spotting the specific functional groups within the molecule.

Okay, let's kick off with what many chemists consider the ultimate answer when you can use it.

X -ray crystallography.

It's like getting the molecule's architectural plans.

The basic idea seems pretty neat.

Shine x -rays at a crystal, the atoms scatter the x -rays, and from that pattern.

You work backwards.

You deduce the exact 3D arrangement of almost every atom.

Bond lengths, bond angles, the whole lot.

It's incredibly precise.

You know how textbooks always draw saturated chains as zigzags?

X -ray shows you that zigzag, like in hexanedioic acid, you can clearly see it.

And the flat carboxylic acid groups, too.

And it's not just for known structures.

It can crack complete unknowns.

There was this coenzyme, methoxidin, used by bacteria, and Amar was struggling with it, but x -ray crystallography figured it out.

Pretty amazing stuff.

But it's not always the magic bullet, is it?

What are the limitations?

The biggest one, absolutely, is the crystal.

You need a good quality, well -ordered crystal.

If your compound's a liquid or it just refuses to crystallize nicely, then x -ray is off the table.

Simple as that.

Another key thing, it usually doesn't see hydrogen atoms very well.

They're just too light, don't scatter x -rays enough, so you have to infer where they are, usually based on the heavier atoms' positions.

Right, which implies it's not always infallible.

Exactly.

There was a famous case, an antibiotic called diazonamide A.

The structure was accepted for about 10 years, based on x -ray data.

But the x -ray couldn't reliably tell an oxygen atom from a nitrogen atom in one specific spot.

They look quite similar to x -rays.

It was only when chemists actually synthesized the proposed structure, and it didn't match the natural compound, that they realized the x -ray assignment was wrong.

The real structure was slightly different.

Wow.

So x -ray gives us that definitive shape, potentially, but you need a crystal, and even then you need to be careful, which really sets the stage for why we need these other methods.

Let's move on to mass spectrometry, MS.

This one's different, right?

It doesn't involve absorbing light or energy in the same way.

That's right.

MS is unique.

It weighs molecules.

Or, more accurately, it weighs charged ions derived from the molecules, not the neutral molecules themselves.

How does it do that?

What's inside the machine?

Well, basically, you get your sample into the gas phase, then you ionize it, turn the neutral molecules into charged particles.

Then you use electric or magnetic fields to separate these ions based on their mass to charge ratio, mildews.

Finally, a detector counts them.

A very common way to ionize is electron impact, EI.

It's kind of energetic.

Imagine firing high -energy electrons at your molecules.

Boom.

One electron gets knocked off, usually a weakly held one.

What you're left with is a positive ion, but also with an unpaired electron, a radical We call this the molecular ion, M plus U, and crucially, the heaviest ion you normally see in an EI spectrum is this molecular ion.

It tells you the molecule's mass.

Okay, like the honeybee alarm pheromone example, heptan 2 -1, molecular weight 114.

Exactly.

You'd see a peak at military 114 in its EI spectrum, representing that M plus ion.

But you said EI is energetic.

Does that cause problems?

It can, yes.

Because it hits the molecule so hard, it often causes it to break apart, to fragment.

Sometimes that fragmentation pattern is useful information in itself, but if your molecule is fragile, you might not even see the molecular ion peak at all.

That's why we have gentler methods like chemical ionization, CI, or electrospray ionization, ESI.

These are much softer.

They usually add a proton, giving you an M plus H plus ion, so M plus 1.

Or sometimes they'll pick up a sodium ion, giving M plus an A plus A, which is M plus 23.

Ah, like the electrospray spectrum for heptan 2 -1 you mentioned, showing a peak at 137.

That's the M plus an A plus an A.

Precisely.

So these gentler methods are great for just getting the molecular weight without shattering the molecule.

But MS does more than just weigh things.

It can hint at the atoms inside.

Right, the atomic composition.

How does that work?

Is it about isotopes?

Exactly.

It isn't just natural isotopes.

Take bromine.

It exists naturally as almost a 50 .50 mix of bromine -79 and bromine -81.

So any molecule with one bromine atom will show two peaks in the mass spectrum for the molecular ion.

They'll be almost equal in height and separated by two mass units.

See that pattern?

You immediately think bromine.

And chlorine is similar, but different ratio.

Yeah, chlorine is about 75 % chlorine -35 and 25 % chlorine -37.

So one chlorine atom gives two peaks, again separated by two mass units, but this time in roughly a 3 .1 height ratio.

It's another really distinctive signature.

Okay, bromine and chlorine have very clear patterns.

What about other elements, like carbon?

Carbon's interesting.

Almost all carbon is carbon -12, but about 1 .1 % is the heavier isotope, carbon -13.

This means that for every carbon -containing ion peak in the spectrum, there's a tiny little peak, one mass unit higher the m plus one peak, and the height of that tiny m plus one peak relative to the main m peak tells you how many carbon atoms are in that ion.

The ratio is basically 100 to 1 .1 times the number of carbons.

It's incredibly useful for quickly estimating the carbon count.

So you get the weight, you get clues about specific elements like halogens, even the number of carbons, but what if two different formulas add up to the same whole number mass?

Ah, that's where high resolution mass spectrometry, or HRMS, is essential.

Standard MS gives you nominal mass, the nearest whole number, but HRMS measures the mass incredibly accurately, maybe to four or five decimal places.

And different atoms don't weigh exactly whole numbers.

Correct.

Because of isotopes and nuclear binding energy, the exact mass of, say, C7H420 is different from the exact mass of C8H18, even though both have a nominal mass of 114.

So HRMS can measure, for example, 114 .1039 for the B pheromone, and that exact mass uniquely matches C7H420, ruling out all other possibilities.

It gives you the definitive atomic composition.

That's powerful.

And you mentioned a nitrogen rule.

Yes, it's a handy little shortcut that often works, especially with HRMS data.

If your molecule has an odd exact molecular weight, it almost certainly contains an odd number of nitrogen atoms, 135.

If the molecular weight is even, it has an even number of nitrogens, 0, 2, 4.

It's a quick check.

Okay, MS gives mass and composition.

Now, for the structure, the connections,

NMR.

You said this is arguably the most important tool.

For many organic chemists, yes, absolutely.

NMR, nuclear magnetic resonance, it's all about the nuclei of atoms acting like tiny magnets.

Like compass needles.

Sort of, yeah.

A certain nuclei, crucially for us, hydrogen -1 protons and carbon -13, have a quantum property called spin.

This spin generates a tiny magnetic field.

Now put these spinning nuclei in a very strong external magnetic field, like inside an NMR spectrometer.

They can align either with the field, which is a lower energy state, or against the field, which is slightly higher energy.

Okay, two possible states.

Right.

And the energy difference between these two states is tiny, corresponding to radio wave frequencies.

If you zap the sample with a pulse of radio waves at just the right frequency, the nuclei in the lower state absorb that energy and flip up to the higher state they resonate.

When the pulse stops, they relax back down, emitting that energy as a radio signal, which the instrument detects.

That signal tells us about the nucleus.

But if all protons are protons, and all C -13 are C -13, why don't they all resonate at the same frequency?

What causes the different peaks we see in an NMR spectrum?

The chemical shift.

Ah, that's the key.

The chemical shift.

It happens because the nucleus isn't isolated, it's surrounded by electrons.

And these electrons,

circulating in response to the big external magnetic field, create their own tiny magnetic field that opposes the main one.

They shield the nucleus.

Exactly.

They shield it.

So the actual magnetic field experience at the nucleus, the local field, is slightly weaker than the external field.

But, and this is crucial, the amount of shielding depends entirely on the chemical environment of that nucleus.

If a nucleus is near an electronegative atom, like oxygen...

Oxygen pulls electrons away.

Right.

It pulls electron density away, reducing the shielding effect.

We call this deshielding.

So a deshielded nucleus feels a stronger local magnetic field.

A stronger field means a bigger energy gap between the spin states, which means it needs a higher radio frequency to resonate.

So deshielded nuclei resonate at higher frequencies, further downfield in the spectrum.

Okay, let's take an example.

Ethanol.

CH3CH2 -OH.

Perfect example.

The carbon in the CH2 group is directly attached to the oxygen.

Oxygen pulls electrons away, deshielding that carbon, so its signal will appear at a higher frequency, further downfield, compared to the CH3 carbon, which is further away and more shielded.

And this difference is measured in BPM.

Parts per million.

Yes, BPM.

It's a relative scale, comparing the resonance frequency to a standard reference compound, usually tetramethylene TMS.

TMS is defined as zero PPM, and most other signals appear downfield, positive PPM values from it.

So when looking at a 13C NMR spectrum, are there general regions we should pay attention to?

Definitely.

13C NMR spectra are usually spread over about 200 PPM, and you can roughly divide it up.

Zero to about 50 PPM, that's typically where you find your standard saturated carbons, like in alkanes.

They're the most shielded.

Then maybe 50 to 100 PPM, these are often saturated carbons, but now they're attached to an electronegative atom, usual oxygen or nitrogen, deshielding them a bit.

From 100 to 150 PPM, this is the region for unsaturated carbons, think alkanes and aromatic rings.

Their electronic environment is quite different.

And finally, above a 150 PPM, often stretching to over 200 PPM, this is where you find carbonyl carbons, CO.

They're very strongly deshielded because of the double bond and the electronegative oxygen.

They really stand out.

So looking at lactic acid, you'd see peaks in different regions for the methyl, the COH carbon and the CO carbon.

Precisely.

Three distinct signals in characteristic regions.

Symmetry is also key.

Hexenedioic acid has six carbons, but due to symmetry, you only see three unique signals in its 13C NMR.

The carbonyls would be way downfield, heptan -2 -1's carbonyl would be obvious, probably around 208 PPM.

And what about 1H NMR?

Proton NMR.

Similar idea, but different scale.

Same fundamental principle, yes.

Spin 12 nuclei protons in a magnetic field, radio waves causing transitions.

But the chemical shift range for protons is much smaller.

Typically just 0 to 10 PPM, maybe 12 PPM sometimes.

Why so much smaller than carbon?

Mostly because hydrogen only has that one electron directly involved in bonding and it's on the outside of the molecule.

The variations in electronic environment just don't cause such huge shifts in shielding compared to carbon, which is embedded within the electron clouds of its neighbors.

And are there general regions in 1H NMR too?

Broadly speaking, yes.

Protons on saturated carbons usually show up between 0 and 5 PPM.

Protons on unsaturated carbons, like alkenes or aromatics, are more deshielded and appear between about 5 and 10 TPM.

So benzene protons are around 7 .5 PPM, quite downfield.

Cyclohexane protons, fully saturated, are way upfield around 1 .35 PPM.

And again, electronegative atoms cause deshielding the methyl protons next to the oxygen and TBME.

Torpedyl methyl ether are shifted down to about 3 .15 PPM.

And you mentioned 1H NMR gives even more detail.

Oh yes.

The real power of 1H NMR comes from something called spin -spin coupling, which tells you about neighboring protons.

But that's a deep dive for another day.

For now, just knowing the chemical shift regions is very useful.

It sounds like putting the NMR data together is how you really start solving structures.

Absolutely.

Take those C4H10O alcohol isomers.

We talked about N -butanol, isobutanol, t -butanol.

Their 13C NMR spectra alone can distinguish them.

N -butanol.

Four signals, all different, isobutanol.

Three signals, because two methyl groups are identical, kegutanol.

Only two signals, because the three methyl groups are identical, plus the carbon attached to the OH.

The number of signals tells you about the symmetry.

The chemical shifts tell you what type of carbon each signal represents.

It lets you piece the puzzle together.

Okay, so NMR maps the skeleton.

What about the functional groups specifically?

That's where IR comes in.

Exactly.

Infrared spectroscopy is the perfect complement.

It's fantastic for identifying which functional groups are present or absent.

It works on a different principle.

Molecular vibrations.

Vibrations.

Like bonds stretching and bending.

Precisely.

Bonds aren't rigid sticks.

They behave more like springs connecting atoms.

They can stretch, bend, waggle, twist.

IR radiation has just the right energy to excite these vibrations.

When a molecule absorbs IR light at a specific frequency, that corresponds to a particular bond vibration getting excited.

The frequency depends on two main things.

The strength of the bond.

Stronger bonds vibrate faster, at higher frequency.

And the masses of the atoms involved.

Lighter atoms vibrate faster.

Think Hooke's law.

Like masses on springs.

So CH bonds with light hydrogen would vibrate at higher frequencies than CC bonds?

Generally yes.

IR frequencies are usually expressed in wave number, CME1, which is proportional to frequency.

A typical IR spectrum plots absorbance, or transmittance, often shown upside down with peaks pointing down, against wave number.

Usually from about 4000 cm1 down to 600 cm1.

And are there key regions in the IR spectrum to look at?

Like an NMR?

Oh absolutely.

Four main regions are super useful.

Well first, the XH region, roughly 4000 -2500 cm1.

Because hydrogen is so light, bonds to H vibrate at high frequencies,

CH stretches are usually just below or around 3000 cm1, NH stretches are sharper, around 3300 cm1.

If you have an NH2 group, you often see two keeks there.

And the most characteristic is the OH stretch.

From an alcohol or phenol, if it's hydrogen bonded, which it usually is, you get a very broad rounded peak, anywhere from 3500 down to maybe 2900 cm1.

Carboxylic acid OH stretches are even broader, often a huge V -shape overlapping the CH region.

If the OH is not H bonded, like in BHT with its bulky groups, you see a sharp peak around 3600 cm1.

So that broad OH peak is a major clue for alcohols or acids, what's next?

The triple bond region, 2500 -2000 cm1.

This area is often completely flat, empty, so if you do see a peak here, it's highly diagnostic.

It's almost certainly either an alkane on CHE, around 2100 cm1, or nitrile C, around 2250 cm1.

OK, XH triple bonds, what else?

The double bond region, 2500 cm01.

This is probably the single most important region for functional groups.

Why?

Because the carbonyl group, CO, lives here.

Carbonyls give a single, usually very strong sharp peak, somewhere between about 1900 and 1500 cm1.

The exact position tells you what kind of carbonyl it is, ketone, aldehyde, esteramide, etc.

Also in this region, you might see alkane CC stretches, around 1600 cm1, usually weaker, and aromatic ring vibrations, often 2 or 3 bends between 1600 and 1500 cm1.

Why is the CO peak so strong?

Ah, good question.

The intensity of an IR absorption depends on how much the dipole moment of the bond changes during the vibration.

The CO bond is very polar, so stretching it causes a big change in dipole moment, leading to a strong absorption.

Symmetrical bonds, like the CCD and ethene, have little or no dipole change on stretching, so their IR peaks are weak, or even absent.

Makes sense.

And the last region?

A low 1500 cm1 is the single bond region, often called the fingerprint region.

It's usually very complex, full of peaks from CC stretches, CO stretches, often around 1100 cm1, CN stretches, and lots of bending vibrations.

It's generally too messy to interpret bond by bond, but the overall pattern is unique to each specific molecule like a fingerprint.

Useful for matching against known spectra, but less so for deducing functional groups from scratch, apart from maybe spotting a strong CO stretch.

Okay, so x -ray for the blueprint, if possible, mass spec for weight and formula, NMR for the carbon -hydrogen skeleton and connectivity,

and IR for the functional groups.

The real power must come from using them all together.

Exactly, it's like detective work.

Each technique provides different clues, and you need to combine them to solve the mystery structure.

Let's try that example.

The unknown industrial emulsifier C4H11NO.

Okay, walk us through it.

Right.

First, ESMS.

We see an M plus H peak at 90.

That means the molecular weight M is 89.

Odd number.

What does that suggest?

The nitrogen roll.

An odd number of nitrogens.

So, one nitrogen.

Spot on.

Hi -res MS confirms the formula is indeed C4H11NO.

Okay, next, 13C NMR.

We see only three peaks, but the formula has four carbons.

There must be symmetry.

Two of the carbons must be identical in the same chemical environment.

Perfect.

And looking at the chemical shifts, one peak is in the CO region, say 60 -70 ppm, and the other two are standard saturated carbons, maybe 20 -40 ppm.

Now, IR.

What does that tell us?

IR shows a big, broad OH absorption, typical of an alcohol.

And it also shows two sharp peaks, around 3 ,300 -3 ,400 centimeters one.

And H stretches.

Two peaks means NH2, a primary amine.

Exactly.

So, we have an alcohol group, OH, and a primary amine group, NH2.

Since the formula only has one O and one N, these must be the groups.

And the carbons that are attached to can't be the symmetrical pair, so the symmetry must come from the other two carbons.

So, C4H11NO with OH, NH2, and two identical carbons.

What structures fit?

Maybe identical methyl groups.

Good thinking.

Two main possibilities come to mind.

Two -amino -two -methylpropan -1 -ol, structure A, or maybe two -aminobutanol, structure B, if the identical carbons are part of the chain somehow.

No, wait.

Two -amino -two -methylpropan -1 -ol fits best.

It has two identical methyls.

Let's stick with A.

It has a C with two methyls, an NH2, and a CH2 -OH group.

That fits three NMR signals.

Okay, structure A seems plausible.

How do we confirm it?

One HNMR clinches it.

If we look at structure A, we expect signals for the OH proton, the NH2 protons, the two CH3 protons, which should be one signal, and the CH2 protons next to the OH.

And let's say the one HNMR shows only two main signals for protons on carbon.

One deshielded signal around 3 .3 ppm, and another more shielded signal around 1 .1 ppm.

That 3 .3 ppm signal is characteristic of protons on a carbon next to oxygen, the CH2 -OH.

The 1 .1 ppm signal fits the methyl groups.

This pattern perfectly matches structure A, two -amino -two -methylpropan -1 -ol.

See how we used all the data.

That's a brilliant example of the synergy.

It really is like solving a puzzle.

You mentioned double bond equivalents earlier as a shortcut.

Yes, DBEs.

Very handy.

Sometimes called degrees of unsaturation.

It's a quick calculation from the molecular formula that tells you the total number of plus double bonds in the molecule.

A triple bond counts as two DBEs.

How do you calculate it?

For a formula like CNH -HS -Nibra -Z,

you can use a formula like DBE equal NM2 plus N2 plus 1.

Basically, start with the number of carbons, subtract half the hydrogen shalogens, add half the nitrogens, and add one.

Each DBE means one ring or one double bond.

So if you calculate, say, four DBEs for C6H6.

Bendy.

Three double bonds and one ring.

Makes sense.

Exactly.

It's a great first step.

If your formula gives you, say, one DBE, you know you're looking for either one double bond like CC or COO or one ring.

If you get zero DBEs, the molecule must be saturated in acyclic.

If you get four or more, especially if the hydrogen count is low relative to carbon, you should immediately suspect an aromatic ring.

Let's try a final quick case study then.

You react protenol with HBr and ethylene glycol and get product X.

How do we identify X?

Okay, product X.

What clues do we have?

MS shows two peaks at 181 and 179 roughly equal height.

Bromine.

One bromine atom.

Correct.

Hi -res MS gives the formula C5H9B -O2.

Okay.

C5H9B -O2.

Let's calculate the DBE.

Five carbons, nine hydrogens, bromines, two oxygens.

DBE equals 592 plus O2 plus 1 plus 1 equals 5.

4 .5 plus 1 equals 1 .5.

Hang on.

Halogens count like hydrogen.

So DBE means five microline, two plus one, two plus one, two plus one, two plus one equals five plus one equals one.

One DBE.

Perfect.

One DBE means one ring or one double bond.

Now 13C NMR propanel had a CO and CC.

The spectrum of X shows those signals are gone.

Instead, we have four signals.

One looks like a CHBR around 30, 40 ppm.

Two look like carbons next to oxygen, maybe 60, 70 ppm.

And one is really interesting down at 102 .6 ppm.

102 .6 ppm.

It's quite far downfield for a saturated carbon, isn't it?

You said C next to two oxygens earlier.

Exactly.

That chemical shift is highly characteristic of an acetyl or ketal carbon single bonded to two oxygen atoms like COCOC.

Okay.

So maybe a cyclic acetyl formed from the ethylene glycol and the bromine added somehow.

What about the IR?

IR, this is key.

No OH stretch.

No CO stretch.

No CC stretch.

Wow.

So the one DBE must be a ring.

It can't be a double bond.

Precisely.

And the IR probably shows strong CO single bond stretches around 1 ,100 cm1, confirming the atherocetyl linkages.

So putting it all together.

C5H9BrO2, 1 DBE which must be a ring, evidence for a COCOC unit, the acetyl at 102 .6 ppm.

Presence of bromine starting materials were propanol and ethylene glycol.

The structure reveals itself.

It has to be 2 -bromomethyl -1 -nephr -3 -dioxylene.

The ethylene glycol formed a cyclic acetyl with the aldehyde and HBr added across the original double bond.

Fantastic.

It really shows how you systematically combine all the pieces of evidence.

This has been a really insightful look at how chemists actually figure out what they've made or discovered.

It underscores a crucial point.

Structures in organic chemistry aren't just drawn, they are assigned based on solid spectroscopic data.

That evidence is absolutely vital for reliable science.

When you read a paper reporting a new molecule, you'll always find it's spectral data listed MS, NMR, IR.

It's the proof.

It really emphasizes that understanding how we know something is just as important as knowing the fact itself.

Absolutely.

Knowledge is one thing, but understanding the evidence behind it, that's where the real power lies.

This journey through spectroscopy has been fascinating, and it feels like we've just opened the door, especially with NMR.

There's clearly more to explore there.

Oh, definitely.

We haven't even touched on coupling in 1H NMR or more advanced 2D techniques.

Plenty more for future deep dives.

We look forward to that.

A huge thank you for joining us on this deep dive today, and thank you, our listeners, for being part of the Last Minute Lecture family.

Until next time, keep exploring the molecular world.