Follow US:

Practice English Speaking&Listening with: Seeing Things in a Different Light: How X-ray crystallography revealed the structure of everything

Difficulty: 0


On the 30th of April, 1852, Thomas Henry Huxley, one of my

scientific heroes, walked through that door to give his

first Friday evening discourse.

His heart was beating like a sledgehammer, as is mine.


He wrote to his wife later, I now know what it feels like to

be going to be hanged.


His nervousness, I think, was partly the

fault of the audience.

He continued, in his letter, "the audience is a most

peculiar one-- at once the best and worst in London."

Such as tonight.

"The best because you have all the first scientific men--

the worst because you have a great number

of fashionable ladies.

The only plan is to take a profound subject--

and play at battle and shuttle-cock with it--

so as to suit both."

Well, Huxley's outrageous division of the sexes is, of

course, of its time.

We know now, thanks to quantum mechanics, that ladies can be

fashionable and intelligent at the same time.


As can men.


But where I agree with Huxley--

and I agree with him on many other topics, as well--

is in the choice of a profound subject.

And I hope I have chosen a suitably

deep subject for tonight.

What I want to tell you about is the technique of x-ray

crystallography, a technique that has a very famous and

intimate relationship with this place.

And it's a technique we are celebrating the 100th

anniversary of the invention of this method, which has

shown us the world in unprecedented detail.

And we're celebrating the centenary this year.

Now, unfortunately, it doesn't really have the appeal in the

public domain that I would like.

It's hidden from us, and that's partly because the

technique relies on a phenomenon that is not

available to us in our everyday lives.

We don't come across x-ray diffraction, which is the

physical principle of which the method relies.

But, it does speak to a broader subject, which is

humankind's obsession with seeing things that are

normally too small to see.

And that's something that, actually, has grabbed the

human imagination for as long as we have been putting lenses

together to make telescopes and microscopes.

And one of the most famous publications from the 1600s--

it was 1665, and we have an original copy here--

is Micrographia, written by Robert Hooke, who was a

pioneer of microscopy.

And this was a very popular book in its day.

It was a bestseller.

Peeps tweeted about it.

It is the most ingenious book I have ever read.

And the secret of Hooke's success, I think, was that he

included images of things that were familiar--

such as his famous flea--

even if they were largely unseen to most people.

And these astonishing detailed pictures of the horrific flea

delighted and enthralled the people who bought his book.

But, the trouble with microscopy is

that it has got limits.

It's fantastic in its own way, but there are limits in a

number of ways.

One is, that when you're illuminating material with

visible light, you, generally, only see the surface.

You see the light that is reflected off the surface of

the material, as with the flea here.

You can't see inside.

And also, you are limited in the size of the things that

you can see.

What if you want to see the molecules that the

flea is made of?

And what I want to tell you tonight is about the use of

x-rays in order to solve both of those problems.

And I hope, in doing so, to sort of set the ball rolling

in a popularisation of the technique.

Because, now that we're celebrating the centenary, it

is a method that has really

transformed all of the sciences.

It relies on a rather beautiful piece of physics and

mathematics, but it has informed not only those

subjects, but mineralogy, chemistry, and biology.

It has really brought us an acquaintance with the world at

the molecular and anatomic level, which was simply not

appreciated prior to the turn of the 20th century.

So, I'm going to make you work a little bit hard a couple of

times during this, as I go through the theory.

But I am not going to delve into all the gory details--

I will spare you that, ladies and gentleman.

But I do want to tell you about how you use x-rays, and

how you use them in a very particular way.

Now, from the very beginning, it was obvious that x-rays was

a special kind of light, because it

could see through matter.

And here you can see one of the very first--

it wasn't quite a medical x-ray.

I think it was more an experiment that Rontgen, who

discovered x-rays in 1895, did-- not on his own hand,

smartly, but on that of his fashionable wife.

And what you can see is that the skin and soft tissue is

largely transparent, because matter is mostly transparent

or semi-transparent to x-rays.

But the bone is slightly denser.

It's got calcium atoms in it, and so it casts a shadow.

And the ring--

let's give Rontgen a bit of credit and assume that it's

gold, which, with an atomic number of 79, is very


And so it scatters x-rays and casts a darker shadow.

So, what's happening here is that the electrons in the

atoms and molecules of Anna Bertha's hand are scattering

x-rays out of the beam, and so it attenuates

the transmitted beam.

And so you see the shadow.

But rather than being a solid shadow, we see that it's

semi-transparent because the x-rays are very penetrating.

So, let's think about the scattering that's happening

and that explains that x-ray pattern--

that x-ray shadow.

So, x-rays, we now know, is a special kind of light.

It just has a very short wavelength, and it effectively

consists of an oscillating electric and magnetic field.

Here, we're just showing the electric field, and it's about

to hit an electron.

And when an electric field hits an electron--

a charged particle--

it causes that electron to oscillate up and down.

And an oscillating electron will, itself, emit radiation

of the same wavelength, and so it emits x-rays.

And, in fact, almost in all directions are x-rays emitted.

And so, the transmitted beam is attenuated

somewhat because of that.

Most of the x-ray is unaffected, but a small part

of the energy is re-radiated--

scattered, we say--

in all directions.

And if we look at a more realistic representation of an

atom, we can see that what happens is, the straight

through beam is attenuated as the x-ray is partially

scattered by the electrons in the atom.

So, what the x-ray is doing is penetrating and sampling all

of that structure that's inside the atom.

Now, when we take medical x-rays, we're only really

looking at the transmitted beam, and we ignore the

scattered radiation.

But the technique of x-ray crystallography actually

relies on the scattered radiation, because it, in

itself, contains information about the object that's doing

the scattering.

And that is the basis of the technique that I want to tell

you about tonight.

And what this technique can do--

because of the penetrating power of the x-rays, it shows

us the interior structure of things.

And it does so, not just at a fairly anatomical level--

we can actually begin to learn about the structures of

molecules and of atoms.

And that happens because of a very

peculiar property of light--

of any kind of radiation--

when it interacts with matter which is of the similar size

to its own wavelength.

And we can get an understanding of that-- that

process, that phenomenon of diffraction.

So what happens is, when light scatters from an object that

is about the same size as the wavelength, it scatters in

many directions.

And we can get an idea of that by repeating classic optics

experiments that were done in the early part of the 19th

century by a polymath, Thomas Young.

And I don't know how Young managed to do these, because I

think he only had candles at his disposal.

But, fortunately, we have a laser.

So, I would like to just show you the principle of

diffraction that occurs in order to give you an idea of

the strange property of light when it interacts with matter.

So, if I just blank the slides and we could have the lights

down a little bit, we'll be able to see what happens.

So, this is a very cheap laser, and the beauty of laser

light is that it gives us a very fine, very bright, and a

very parallel beam.

And that beam is travelling in a dead straight

line from the laser--

bouncing of the mirror, of course--

to the screen.

So, we see a fine spot.

But, if we then take an object, a very simple object--

and this is just a single slit, so there's a small gap

here, which is of approximately the same size as

the wavelength of light--

we can see--

let me see.

Let me get that right.


So, you have to look a bit closely, but what you see is

we no longer have a simple spot.

And we don't have an image of the slit, either--

the slit is just a rectangular aperture.

But, you see, what has happened is the light is

spreading out at an angle.


Not necessarily very much here, but enough

that you can see it.

And you can see there's a bright band in the middle, and

then flanking it on either side, if my laser holds up--

well, there we go--

then you can see that there are fainter bands on either

side, as well.

And what that tells us is two things.

One is that the light is bent--

it's scattered at an angle by the small object--

and that there's a pattern here that appears to be

related to the particular structure that's doing the


And so, there's a relationship that, if you understand the

rules of optics, we could work backward from that pattern to

figure out that we had a single slit here.

Now let's see what happens when we put two

slits into the beam.

So, these are just two slits, both the same width as the

original slit in the first slide.

And now, we can see that there is a different pattern,

although it is similar to the first one.

So again, there's a central broad band, and its flanked by

fainter bands on either side.

But now, each of those bands is actually split into a

series of dots.

So, again, that tells us that we've changed the structure,

here, that's doing the scattering, and that changes

the diffraction pattern that we see on the wall.

But, because the structure here is simply a duplicate of

the first structure, the overall pattern-- the

envelope, as it were-- of the intensity variation changes.

It's just that we see different points of light.

And if we go from two slits, now, to six, then we see that,

again, we get a similar overall pattern--

a, sort of, central bright than flanked by fainter bands

on either side.

And, again, now they are divided into dots.

And the pattern of dots is exactly the same as with the

two slits, because the spacing between the six slits is the

same as the spacing between the two slits.

And that's a point I want you to keep in mind because we'll

come back to talk about that when we talk about crystals

and what crystals do.

Because a crystal is, simply, a repeating form of matter.

You have a structure that is repeated, and we will be using

that to probe the structure of matter.

But the optics shows us how that principle works.

So, that's my diffraction, but--

so, I've mentioned crystals there, and I showed you that--

with a repeating pattern of the slits--

was important for getting an interesting diffraction

pattern on the screen, that we could then start to think

about it and interpreting.

Although I haven't explained to you, yet, how you do that.

But, before I do that, we want to, then, actually grow some

crystals of our own.

So, if Jayshan, my assistant, is going to come here.

And so Jayshan--

he's a novice crystallographer.

And so, what Jayshan is doing here is, we have a solution of

a protein--

a purified protein-- called lysozyme, which some of you

may be familiar with.

OK, we have pre-filled these small wells with a solution

that's got a high concentration of sodium

chloride and an organic solvent, called polyethylene

glycol, or PEG.

And Jayshan is currently mixing some of that solution

in with a drop of protein in this well, and he's then going

to seal the chamber.

And what's going to happen over the

course of the lecture--

we hope--

is that, because we have a high concentration here, you

have a lower vapour pressure above the reservoir.

And so water gradually moves through the vapor phase, out

of the drop and into the reservoir.

And what that will do over a course of minutes, that will

reduce the volume of the drop, and it will increase the

concentration of protein inside the drop to a point

where, we hope, it starts to precipitate out of solution.

And when a protein precipitates out of solution,

what that means is that the molecules are starting to

stick together.

They interact with one another rather than interact with the

water molecules.

And what we hope is that they will aggregate together in a

very regular array to give us a protein crystal.

So, we are almost done with that.

I did say he was a novice.

I was going to do this myself.

I thought I'd be too nervous, but my hand looks OK.

The trouble is that's the hand I pipet with.

I believe that is a Tommy Cooper joke, just to give

proper attribution.

So, we are going to put that on one side.

We want to take it out of the bright lights in case the

temperature perturbs it, but we will come back to that

experiment later in the lecture.

But, I'm getting a little bit ahead of myself with crystals.

I've told you about x-rays and diffraction.

Let's think about how the two were finally brought together.

So, as I said, x-rays were discovered in 1895 by Rontgen.

And they were quite a puzzle at the beginning, although we

now know that they are a type of electromagnetic radiation,

and so they are waves.

At the time, there was a great debate-- are they waves or are

they particles?

It really wasn't known, and many people spent a lot of

time trying to investigate their properties in order to

work it out.

And, in Germany then, a scientist called Max Von Laue

thought that he would be able to crack the problem.

Because he reasoned that inside crystals, and in a

crystal of copper sulphate-- this is a beautiful blue

crystal, here--

it was well understood, because of the regular faces

that one sees in crystalline solids--

it was already presumed that the atoms or molecules within

them would be lined up in regular arrays.

And Laue figured that the wavelength of x-rays was

probably about the same size as the spacing between the

atoms in crystals like copper sulphate.

And so, he persuaded to Walter Friedrich and Paul Knipping to

do the experiment for him.

So, I guess they were the Jayshan's of their day.

And, what he did was, he took a Crookes tube, which

generates x-rays, and put it through fine slits in order to

get a pencil beam, as they called it then.

Placed a crystal here, in their experiment, and then

placed a photographic plate at the far side.

And they exposed that for several hours, and then

developed it on tenterhooks in the dark room, and then

produced one of the most remarkable

images of the 20th century.

Now, I know it doesn't look like much, but it is a truly

significant scientific result.

It does look like a horrible smudge--

maybe it's a Rorschach diagram, I'm not quite sure--

but what you can see are two things.

One, there is a diffraction pattern there.

It's not a very regular diffraction pattern, not

immediately apparent.

But you can see that the beam has been split into different

rays, and so it's scattering.

So there's a ray here and a ray here, and these are spread

out from the middle.

So the audience, here, is where the x-ray beam came

from, and this is the pattern.

So, Laue's hypothesis was supported, so it did suggest

that x-rays are waves and that crystals could be used as a

diffraction grating for them.

But-- as well as showing that they were waves--

because you have a pattern here, what that tells you is

that the structure within the crystal--

because the beam has passed straight through and been

scattered off at angles--

the structure of the crystal has imprinted itself, somehow,

on the pattern of scattering.

And so, what you have in this pattern--

and when they moved the photographic plate back a bit,

they got a slightly better resolved

separation of the spots--

but there's information in here about the structure of

the crystal that they were analysing.

Now, this wasn't the most regular pattern.

It turns out, there were good reasons, initially, for

choosing copper sulphate, but--

although it is a nice, regular array--

the atoms are not arranged, in copper sulphate, in a very

symmetric pattern.

So they then chose a crystal of zinc blende,

which is zinc sulphide.

And from the appearance of these beautiful crystals, you

can tell that it does look like the atoms are lined up in

a cubic type of formation.

And that certainly turns out to be the case.

And if you take one of these crystals and you orient it

properly in the beam, then you get a much nicer pattern.

And you can see, actually, that the pattern of scattering

of the x-rays has a four-fold symmetry.

You can see this makes a sort of square that is hinting at

four-fold symmetry, which you would expect for a cube,

inside the crystal.

But, Laue tried to analyse this pattern and he was mostly

able to solve it, but he wasn't quite able to

get the whole way.

And this is in 1912.

So, it was a major breakthrough.

It was big news in Europe, but Laue himself wasn't able to

solve the structure, wasn't able to figure out exactly how

it is that you use x-ray diffraction information in

order to work out structures.

That fell to the father and son team of William and

Lawrence Bragg.

Now, William was already, at that time, a renowned

physicist and had been working on the problem of x-rays up to

that point for the past 10 or 15 years, and was a world

expert on that.

And so, a Norwegian colleague of his, Lars Vegard, wrote to

him, because he'd been to Munich and

he'd talked to Laue.

And he sent a letter saying, "Recently, however, certain

new curious properties of x-rays have been discovered by

Dr. Laue in Munich." And the letter was a long and detailed

account of the experiments, and Laue was even kind enough

to give Vegard a photograph, which he then knew he was

going to send the Bragg's.

So, this was a very nice example of scientific

international co-operation.

And this happened in the summer of 1912, and Lawrence,

William's son, had just completed his second degree--

a degree in physics-- in Cambridge.

He'd earlier got a degree in mathematics when they were

living in Adelaide in Australia.

And it was Lawrence, really, who was able, ultimately, to

crack the problem of figuring out how there is a

relationship between the scattering pattern of the

x-rays and the structure of inside the crystal.

And it was fortunate for him, in a way.

There was a sort of coming together of events.

He'd just finished his physics degree, so he had been

learning about electromagnetic radiation, he'd been learning

about the symmetry of crystals.

And he also was blessed with a father who was a world expert

in x-rays and who had, then, received this letter.

They hadn't actually seen Laue's paper.

And Lawrence realised that the key to the problem was

realising that what's happening is that the x-rays

are reflecting off planes of atoms in the crystal.

And so, I'm going to make you work hard here by explaining

to you a little bit of the theory.

So, we are all familiar--

at least those of us who wash regularly--

with the laws of reflection.

So, you know that the angle of incidence equals the angle of

reflection for light impinging on a

silvered or polished surface.

Now, that law of reflection applies equally-- although I

won't explain the details--

even if that surface is discontinuous--

say, for example, it was made up of a set of atoms.

But remember, if we're now thinking about what x-rays do

to atoms, matter is mostly transparent to x-rays.

And so, a large fraction of the energy will

actually pass through--

certainly the top layer of atoms.

Only a small fraction of it will be reflected.

Most of the energy-- more than 99%--

goes straight through.

And so, that will interact, then, with the layers of atoms

below that in the crystal.

And we have regular arrays of atoms-- many, many of them.

And each layer will then, itself, reflect a tiny

fraction of the incident x-rays.

And what Lawrence realised was that what he needed to do was,

to work out how much the beam is scattered in this

direction, he needed to add up all of these rays.

Now, this diagram is getting a little bit complicated, but

Lawrence had a gift, I think, for simplifying things.

So, let's just think about one of the rays going through, and

you get a partial reflection of every single layer within

the crystal, from this law of reflection.

Now, we also have to remember that what's happening here is

an electromagnetic phenomenon, so the x-rays are waves.

And so, we have to think about what the waves are doing.

And to make it even simpler mathematically, we can boil

that down to just two waves, although let me just tell

you-- so here, the scattered rays are what we call in

phase, and what that means is that the peaks are lined up.

And so, all the peaks are lined up and all the troughs

are lined up.

And when that happens, when you add those waves together,

you get a very big wave--

a wave with a very big amplitude.

It oscillates massively.

So, that's called constructive interference.

And so, that's the best way that waves can add up,

although there are other relationships,

too, as we'll see.

But let's plough on and see how these relationships arise.

So, if we just think about by what's happening in the

reflections of adjacent rays-- so the top

one and the one below.

So, we have the incident ray-- comes here and goes away.

And then the lower ray hits the bottom layer and is

reflected off the lower layer.

So let me just get that.

So what you see is that the lower ray has to

travel a bit further.


So it sort of falls out of step with the upper ray,

because it has a shorter path to go around.

It's like running in the outside lane on an athletics

track, as it were.

But, as long as this extra distance travelled--

which is indicated by the two black arrows on my diagram--

is equal to one wavelength, or two wavelengths, or three

wavelengths, or any whole number of wavelengths, then

the waves will be back in step.

And so, they will add up, and you will get appreciable

scattering in this direction.

So, Lawrence Bragg worked out, by simple trigonometry, that

this extra distance from the spacing d, and when you're

reflecting at an angle theta, the path difference is equal

to 2d sine theta.

And if that path difference--

2d sine theta-- is equal to a whole number of wavelengths--

so, the wavelength is simply the distance between two

adjacent peaks--

then you will get appreciable scattering in this direction.

And that is Bragg's Law.

Now, if you look at a slightly different angle--

and here we've just changed the angle of incidence--

and in this particular case-- the way that I've chosen it--

the path difference is such that the lower wavelength is

out of step by half a wavelength, or a whole odd

number of half wavelengths.

So, in this case, the peak of the top ray lines up with the

trough of the bottom ray.

And that means, in this case, that these two rays cancel one

another out, so you get no scattering.

Now, if you think about it, you'll have the same

relationship for the next pair of layers down, and then the

next pair after that, and so you get no

scattering in that direction.

And so, what Bragg's Law tells you is which directions you're

going to get scattering in, depending on the angle that

you're looking at and the spacing--

which is an indication of the structure--

within the crystal.

Now, you might think, well, OK, he's just shown us two

particular special directions there.

But, actually, Bragg's Law is a very severe law, and it only

allows scattering if it is true.

So, let me show you how that works out by a slightly more

sophisticated example, and I'm going to come to the front so

I can see this with you, as well.

So, let's think about a case where it's coming in at an

angle, and the difference between adjacent layers is

such that the path difference is only 1.01 wavelengths.

It's a percent out.

So, this ray--

this wave-- is almost exactly in phase with the lower one.

Near as dammit.

And if you, then, think about the next ray up, that is twice

as far away from the first ray.

And so, the path difference is doubled.

And in this case, the path difference is 2.02 lambda,

which is, again, almost in phase with the lower one.

So you kind of think, well, this is going to add up--

there's going to be quite an appreciable amount.

And the same for the third one-- it's only 3% out.

It's not really too bad.

But, by the time you get up to the 51st layer--

and remember that in any given crystal there will be

thousands upon thousands of layers, if not

millions upon millions--

when you get to the 51st, the path difference is 50.5


And so, this ray is half a wavelength out of step with

the one from layer one, and so, those two will cancel out.

The diagram is a little bit misleading, but they are close

enough in space that you do get interference here.

And so, the 51st layer will cancel with the first one.

What that means also, then, is that the scattering from 52 is

going to be half a step out with the

scattering from layer two.

And so on, up the stack of layers in the crystal.

So, unless Bragg's Law is satisfied--

and that is, bang on, one, two, three, or four


then you do not get any scattering.

And so, Bragg's Law places a severe restriction on the

scattering of the planes.

But it also tells you interesting information about

the spacing between those planes.

And also--

we've just looked at one horizontal set of planes in

the crystal--

but there are many other ways of looking at crystals.

These are the horizontal planes, but you can equally

imagine that the atoms line up in a completely different way.

And so, we have an angled set of planes here.

But you can look at the same thing again, and you have

another set of angled planes, each with different spacings.

And what Lawrence realised was the spots in a diffraction

pattern are simply all the various different reflections

that are allowed by his law coming off the

interior of the crystal.

And we can maybe get an idea of that just if we look in

three dimensions.

So, here is a atomic lattice, but if we sort of tilt it in

three dimensions, you can see that we see a horizontal set

of planes, there, from the way that the atoms line up.

But we can rotate it again, and here we have another set

of planes which are sloping down from left to right.

And then one final one--

rotate another direction, and here you can see there's

another set of planes sloping from right to left.

And so, all of those sets of planes are inside the crystal.

And when x-rays come in to the crystal, they are bouncing off

those sets of planes in all different directions.

And so, when he looked at a diffraction pattern-- oh, let

me get back to the slides--

when he looked at a diffraction pattern, Lawrence

realised that what he was seeing was each of these rays

was one of the reflections off one set of planes.

And so, it was telling him-- and the angle of the ray,

which you could measure--

was telling him about the angle and the separation of

the planes in the crystal.

And he realised that by, if he could figure out where all the

angled planes were, then the atoms would be lying at the

intersections of all those planes.

And so, he would then be able to work out, once he'd solved

that puzzle--

he would be able to work out where the atoms were.

And so, he was the first, in that case then, to solve the

structure of the atomic arrangement

with inside a crystal.

And this is from zinc blende, and this is Lawrence's

interpretation of the pattern, or his prediction of what the

diffraction would look like, once he had

figured out the structure.

And as you can see, there's a very good correspondence.

And he did this just in the few months after the summer of

1912, so this was by December of that year.

And he'd gone back to Cambridge, because he was

working with J.J. Thompson at the time.

But, he was in regular

correspondence with his father.

"Dear Dad," he writes, and is, sort of, describing the


And it's a nice, sort of, chatty letter.

He's sending him some photographs, but some very bad

prints from the photos.

They're not very good.

But you can see there's a very, sort of, nice statement

about his excitement at the achievement.

So, if you can read the handwriting-- it's better than

mine, I have to say--

"Larry's thing was equivalent to reflection but of course he

didn't see it, and it's great fun getting it straight off,

isn't it?" So, it's the young Bragg--

he was 22, 23 at this time, and you just get a real sense

of his excitement.

And it was a real puzzle-solving exercise for

him, and he was jolly pleased with the result, and quite

right, too.

So, they went on.

Soon, the Bragg's father and son team are working together

analysing lots of crystal structures.

This is one of their own early pictures.

This is from a crystal of potassium chloride, which is a

close relation of sodium chloride.

And again, Lawrence was able to interpret the diffraction

pattern, and this his own prediction from having worked

out the structure.

You can see that he's annotated it, and here--

sorry if--

so, you can see that he has annotated it.

And then, they also analysed sodium chloride, which

actually has a similar atomic arrangement.

But you can see that there are small

differences in the pattern.

You can see, you get many of the spots in the same places,

but some of them change in intensity and darkness.

And this was an early hint that it's not just the

positions of the spots that are important, but the

intensities are as well.

Now, initially they had mostly just been working on the

positions of the spots, which told them about the angles of

the planes.

And it was by working out the angles of the planes that they

could figure out where the atoms were.

But this was an early indication of a more

sophisticated analysis that was to follow it.

So, this was a structure of sodium chloride that they

worked out.

They published it in 1913, and even then-- although they

worked hard on it-- he wasn't entirely sure of himself.

It was on this rather slender and indirect evidence that I

assigned the structure, in a paper read to the Royal

Society in 1913.

Fortunately, few further investigation established its


Now, it was a major breakthrough.

This was the first time that people had seen the interior

of the matter, had seen the atomic arrangements inside a

piece of crystalline matter.

However, it didn't please everybody, and it took the

chemists actually, in particular, a long time to get

on board with crystallography--

or some of them, at least.

So, Bragg was writing, as late as 1927, "In sodium chloride

there appears to be no molecules represented by

sodium chloride." The chemists had expected to see molecules,

of an atom of sodium and an atom of

chloride stuck together.

"The equality of number of sodium and chloride atoms is

arrived at by a chessboard pattern of these atoms.

It as a result of geometry and not a pairing of the atoms."

Now, a British chemist, by the name of Henry Armstrong, took

exception to this.

He got rather cross, and he wrote in the pages of Nature,

"This statement is absurd to the nth degree,

not chemical cricket.

Chemistry is neither chess nor geometry, whatever x-ray

physics may be." You can just hear the disdain in this

voice. "It's time that chemists took charge of

chemistry once more. " So, he was a little bit out of step,

perhaps, with many of his colleagues, because many

chemists did adopt crystallography.

But it didn't phase the Bragg's .

They carried on at working with crystallography.

An early structure that they solved the following year was

the structure of diamond.

And that was followed, and again, 10 years later, by the

structure of graphite.

Now, both of these are forms of carbon, and this shows how

understanding the atomic structure helps us to

understand the material properties of these.

So, in diamond, carbon is bonded in a

three-dimensional pattern.

We have a tetrahedral arrangement, and you have

strong bonds in all three directions.

Where as in graphite, again, a pure form of carbon, the atoms

are arrayed two-dimensional hexagonal nets.

And these nets can slide past one another, and that is why

graphite is used as the lead in your pencil, and is very

soft and leaves a smudge on a paper.

So, understanding the intrinsic structure helps us

to understand exactly what the material

properties of these are.

So, we're getting new insights into material

science from this.

And the Bragg's kept going further and further into this.

They solved more and more structures, mainly, initially,

working on types of material that only

naturally occur as crystal.

So iron pyrite, calcite, and quartz.

And, I agree with him when he wrote-- and, again, this is

early June, 1914--

"we are scarcely guilty of over-statement if we say that

Laue's experiment"-- and it was good of

him to credit Laue--

"has led to the development of a new science." And I think

that's absolutely true.

Now, until that point, they had only been looking at

relatively simple structures of types of matter that

naturally occur as crystals, such as the beautiful crystal

of rock salt that we have here.

And, they had been largely relying simply on measuring

the positions of the spots in the diffraction pattern, in

order to, then, solve the puzzle of how to work out the

internal structure.

But again, early on--

1915, this is--

William Bragg give a lecture at the Royal Society and

identified the opportunities for advancing the technique

for improving it and applying it to more complicated


So, in this simple case, we'd be considering--

the consideration of the crystal symmetry--

though unable themselves to determine the crystal

structure, comes so near to doing so that a few plain

hints given by the new methods--

that is, the positions of the spot in

the diffraction pattern--

have been sufficient for the completion of the task.

The exact positions of the atoms are then known, so they

could work it out.

This is not the case with more complicated crystals, and he

realised that a more sophisticated

mathematical approach--

he realised, even early in 1915, that Fourier methods

needed to be applied in order to bring in the information

that was in the intensities of the spots in the diffraction

pattern, so that they could start to analyse more

complicated things.

Now, I'm not going to go through Fourier Theory with

you here, tonight, in gory detail, but I do want to give

you a type of graphical explanation for it.

So, let's think about a more complicated molecule.

So, this is a complicated molecule.

I hope you'll agree it's a little bit more complicated

than the two atoms of sodium chloride.

This is in fact a protein--

doesn't really matter and what it is-- but you can see that

it's a complex set of bonded atoms.

And what I'm showing here, in this blue mesh, is the

electron density.

So we see exactly where the electrons are.

So, this is a representation.

So, let's think about how x-rays interact with a

molecule like this.

So, an x-ray comes in from the side--

and here it comes--

and, as we saw before with the electron and with a single

atom, the x-ray illuminates the whole of the molecule, and

we get scattering in every direction from every part of

the molecule.


And what you'll notice here is that some of these scattered

rays are beefier than others.

So, this is quite a strong one, and it's coming off in

this particular direction.

We have quite a strong oscillation here.

Whereas is this one is quite weedy, and what that means is

that this is scattering from a point where there's quite high

electron density-- strong electron density--

whereas here is a part of the structure which has weak

electron density.

So, although the scattering is a bit of a mess because it's

in all directions, what each scattering is doing-- it's

carrying off a little bit of information about the electron

density at the point where it was scattered from.

Now, we can't do mathematical analysis on things like this.

We like to break down the problem and

take it step by step.

So, let's simplify it, now, just by considering all the

scattered rays in one particular direction.

There will be thousands of them--I have only had

time to draw four.

So, again, the amplitude--

so, the size of the oscillation-- varies according

to position, and that will vary in different positions r,

indicated by a vector in the structure.

Now, each of these waves is heading off in a particular

direction and can be represented mathematically by

this function.

Now, this function looks horrendous OK,

I agree with you.

But, it's actually not very difficult.

So, as we said, the amplitude depends on the electron

density, and this function here-- rho of r--

tells you, what is the electron density at position r

in the molecule.

So, that's a measure of the electron density.

If that's big-- that number--

then you'll get a big amplitude.

And this dxdydz is just the little volume of electron

density that we have at this point.

This exponential function looks a bit odd, but that's

just encoding the fact that this is a wave--

it tells us about the wavelength.

It also depends on s, which is a number that just varies with

the angle at which we are thinking about.

So, we're just thinking arbitrarily about this one

angle, but we generate the mathematical methods just for

that direction.

But it also tells us about the phase, and the phase of the

wave depends on r.

And so, the phase has got information about the position

of the electrons doing the scattering, as well.

And the phase is simply, where is the position of the first

peak relative to some arbitrary origin.

So, to calculate the total scattering in this direction,

we have to add these up.

And when we do addition of funny terms like this, we have

to integrate--

so again, it looks horrible--

but all we're doing is just adding up the waves.


And when we add up the waves, we get one particular wave

that travels in that direction.

And we can think about the whole contribution from the

whole molecule emerging from some arbitrarily chosen point

in the middle, and then we have a wave--

which we describe is this function, f of S. But this is

just an expression.

This means, a wave scattered in the direction associated

with f, which we also know as theta.

So, that's all the addition.

So, that's that particular direction, but

we would also then--

to calculate, we could sum up, then think about, all the

waves scattered by all parts of the

molecule in another direction.

But, of course, we've got to think about all possible

directions because that will give us the most possible


So, if we then were to, sort of, turn the detector around

to face us and see what happens, then you would-- you

might-- get something that looked a bit like this.

So this doesn't look very regular.

It looks a bit messy.

But there's a pattern of, sort of, dark and light areas.

We've got strong scattering here, medium strength

scattering here, quite weak scattering here, quite weak

scattering there.

But this is a very definite pattern, and this is the

diffraction pattern-- it's just like the optical pattern

that we saw from the laser at the beginning.

And it's got information in it about the structure of the

molecule because it's the x-rays that have come--

scattered from all points within the molecule, in its

three-dimensional interior.

And this mathematical formula, that we've worked out, is

actually known as the Fourier transform, after the French

mathematician Jean Baptiste Joseph Fourier.

What's interesting about Fourier is that he lived and

died before the x-ray was even discovered, but his

mathematical methods have turned out to have very

general use in physics.

He won a prize from the French Academy of Sciences, but they

were a bit sniffy about it.

They gave him the prize, but they did note in the citation,

"the manner at which the author arrives at his

equations is not without difficulties, and that his

analysis for integrating them still leaves something to be

desired." But there you go, that's the French for you.

Lord Kelvin, an Ulsterman, was right on the money when he

said it is "one of the most beautiful results of modern

analysis." Now, the beauty of Fourier's analysis when it's

applied to x-ray crystallography is that we can

work out what the scattering should look like from the

electron density, but if you can calculate the Fourier

transform mathematically, you can actually then, also,

simply do the inverse Fourier transform.

So, we have, basically, rearranged the equation.

You're allowed to do this, according to the


So, if you measure all the scattering, you can work out

from the maths what the electron density is.

And this is basically a mathematical description of

the shape of the molecule.

And so, this is entire mathematical basis of x-ray

crystallography that we use, and still use,

in the modern era.

It's just the maths is all now done inside a

computer, thank goodness.

This mathematical technique would allow us to work out the

structure of a single molecule.

However, we cannot work with single molecules.

They are really small, so they're very hard to pick up.

And even if you could pick it up and put in an x-ray beam,

it was scatter so few photons-- so few x-rays--

that you wouldn't be able to measure it.

So, the trick to get around that is to try and crystallise

your molecule.

And so, you take your protein, or whatever compound you're

interested in analysing, and you try to grow

a crystal of it.

It may not occur naturally, but chemistry has shown us

that the final purification step is often crystallisation.

So many compounds can crystallise, and we now know

that even many proteins can crystallise, as well.

And so, if you get a crystal, you'll see that it behaves

exactly like a crystal of sodium chloride, or of salt.

All you have is a regular ray-- not of atoms this time,

but of molecules.

But you can identify planes, just in the same way as we did

for sodium chloride in zinc blende.

And so Bragg's Law still applies, and so there are only

certain directions in which you will get diffraction.

But, at least, as we saw with the slits when we put in six

slits rather than two, we got much brighter diffraction

because we were allowing more light through.

And so, the crystal basically gives us an amplified signal

that we can interpret.

So, I wonder, now, can we have a look and see how our

crystals of lysozyme are doing?

By the magic of-- oh, and it looks quite nice.

So, this was a clear drop earlier, but you can see,

here, there's a sort of grittiness to it.

But this looks like a whole cluster of little jewels.

And so, these are crystals of lysozyme--

a protein--

that have grown in the last 20 minutes, or so,

that I've been talking.

They're a little bit small.

But, by modern standards, they're perfectly adequate,

perfectly serviceable for doing x-ray diffraction.

So, as long as you can crystallise it, you can solve

the structure of it.

And that was what the Bragg's realised,

back as early as 1915.

So, the application of the Fourier methods allowed


allowed the Bragg's, initially--

to use the positional information, and the fact that

the intensity of the spots varied in different

directions, in order to calculate the

electron density map.

So, they applied Fourier methods, and this is one of

the very first, published by Lawrence Bragg.

This is from 1929, from diopside--

still a fairly simple structure, but much more

complicated than sodium chloride.

And so, this is a section through the

electron density map.

You can see the contours.

So, here's strong electron density here.

This is kind of medium.

And here is the structure that Bragg built into that electron

density map and solved for diopside.

And so, through the '20s and '30s they showed that they

could apply the technique to more and

more complex molecules.

And the chemists didn't all listen to Armstrong.

And this is beautiful work done by one of Britain's, sort

of, most celebrated female scientist, Dorothy Hodgkin--

or Crowfoot, as she was before her marriage.

This is the structure of penicillin that was solved

during World War II and published shortly afterwards.

And this, she published in the late 1950s, is the structure

of vitamin B12--

a massive molecule.

And it was the biggest molecule to have been solved

at that time.

So, a single molecule has over 100 atoms.

And so, the technique moved on in power, thanks to the

application of the Fourier method, which had been

pioneered by the Bragg's.

And crystallography is now, simply, an

embedded part of chemistry.

So, as well as telling us about material science, it's

now a regular routine tool of chemistry.

The database has over half a million crystal structures in

it, and 40,000 new structures are added every single year.

So much for chemistry, which isn't really my subject.

What about biology?

So, again, this started relatively early-- so it was

in the '20s and '30s that people started to think about

how you could apply x-ray methods to look

at biological problems.

Biological proteins are much bigger, in general, and so

it's much more challenging and demanding.

But again, Bragg-- father and son, actually, were both very

instrumental in inspiring and guiding people to

tackle these problems.

One of the first to get involved was a chap called

Bill Astbury, who worked here at the Royal Institution as

part of William Bragg's group in the '20s and

then moved to Leeds.

And here we see the work of one of his Ph.D. Students from

1937 or '38, Florence Bell.

And this is actually one of the very first x-ray

diffraction patterns of DNA-- of nucleic acid.

Now, you don't have a pattern of spots because this is not

exactly a crystal that they're analysing here.

This is a fibre.

But DNA has a very regular structure so it is


And you can see, even in this early pattern--

this is 1938--

the typical, sort of, X pattern that is characteristic

of the double helix that was eventually to emerge.

Now, the whole story of the work on the structure of DNA

in itself is convoluted and tortuous, and deserves a whole

lecture in itself.

What I want to focus on tonight is the crystal side of

biological crystallography.

And that was kind of kicked off by this chap, J.D. Bernal,

an Irishman from Tipperary, who worked--

one of his first students was Dorothy

Hodgkin, or Dorothy Crowfoot.

And Bernal-- he started off as a theorist.

He wasn't very good with this hands, initially.

But under William Bragg's tutelage, here at the Royal

Institution, he eventually mastered the technique and

became a crystallographer of some renown.

And so, people would just send him crystals.

And so, a friend of his was travelling in Sweden in

Svedberg's lab.

They had accidentally grown crystals of pepsin, and their

friend said, I know someone who would give his eyes for

those crystals.

And so, he was allowed to take them away--

carried them in his coat pocket back to London--

or, sorry, back to Cambridge, where Bernal was

working at the time.

And they put them into the beam, and they produced the

very first x-ray diffraction pattern

from a protein crystal.

It must have been quite a moment.

Unfortunately, the photograph that they took is lost.

It was probably destroyed in a bomb during World War II.

However, it probably looked something like this.

So, this is a diffraction pattern from a crystal of

haemoglobin, which was taken by Max Perutz who was also a

student of Bernal's, just a few years later.

But, you see a fairly similar--

must have been very similar to the pattern from pepsin.

And you can see many, many spots in here--

you get many more spots because the

molecules are much bigger.

Now, at that point, they couldn't really analyse the


they couldn't work it out because it was too


It was even beyond the Fourier methods that

they had at the time.

But they realised--

and this is 1934 that it was initially published by

Crowfoot and Bernal, so Dorothy and-- well, Sage, as

he was known, because he was such an intelligent and

well-read person.

Everybody called him Sage.

And there was two things they realised from this.

One was that, because they had regular diffraction, it meant

that the protein molecule is of a perfectly definite kind.

That meant that the protein molecule

had a definite structure.

Before that, until that point, there had been a lot of debate

on this point.

It was thought that it was like a colloid-- a rather,

sort of, loose association of peptides.

But here they showed, because it diffracts, then it is a

definite structure.

And they further realised that they now had, through this

x-ray method, the means of really getting to grips with

what proteins looked like and what they could do.

And, although it was going to take quite some time before

they could realise this, they knew-- and certainly, Astbury

and Bernal knew--

that they were on the cusp of, I think, a major breakthrough.

And there's a very nice quotation in a letter from

Astbury to Bernal--

they wear pals--

where Astbury says, "If you and I do not make the most

biological crystallography, we should have our respective

bottoms kicked."

While I don't think they deserve to have their bottoms

kicked, they are, unfortunately, not as

well-known in the scientific world as they deserve to be.

Their names are well-known inside the

crystallographic community.

That's partly, I think, because they didn't,

themselves, solve any landmark structures.

But they both did early work and inspired other people to

go on to work on structures that were, themselves,

landmark achievements.

Astbury's initial work helped to inspire Maurice Wilkins to

get involved in analysing DNA, and Wilkins was very

instrumental in pursuing the project to fruition.


he had mentored Hodgkin and, also, people like Max Perutz.

But it took until 1959 before the very first protein

structure was solved, and what a moment that must've been,

when this model appeared.

I won't offend the sensibilities of the find,

ladies and gentlemen of London, by telling you what I

think it looks like this.

But, the disappointment in Perutz's voice is palpable--

"Could the search for ultimate truth really have revealed so

hideous and visceral an object?"

Part of the disappointment was because the crystals they

initially had weren't very good, and they didn't scatter

to high angle.

And so, that limited the information that they could

get, and so the model really just shows the fold of the

polypeptide chain--

that reveals the structure.

But you don't really get any biological insight from a

model like this.

And it certainly-- this came six years after the structure

of DNA had been solved--

well actually, DNA had been solved on the basis of much

sparser information, but it produced this beautiful,

double-helical molecule.

It was just elegant in its simplicity, and it immediately

suggested the mechanism for how genetic information is

transmitted from copy to copy.

However, within a couple of years, the crystals improved

and their analysis improved.

And soon, they had a model of the protein structure that

really did have all the atoms in it and was starting to give

us real biological and mechanistic insights.

So, this is myoglobin.

This is a protein that comes from a sperm whale.

It is an oxygen storage molecule, and so it is the

molecule in the muscles of the sperm whale that allows this

incredible beast to hold its breath for a very long time.

So, the initial myoglobin structure was

solved by John Kendrew--

that was done in the Cavendish lab at Cambridge, again, under

the watchful eye of Lawrence Bragg, who was director of it

at the time.

And Perutz himself, working in the same lab, produced the

structure of haemoglobin a few years later-- this is in the

early '60s.

And this is a protein that's found in the

blood of all mammals.

It's the oxygen transporter, so it carries oxygen in your

blood from the lungs, all the way through your tissues, and

back into the lungs.

And immediately this was just the second molecule structure

to be solved--

but what was remarkable about this was that when they

overlaid the structure of myoglobin on it, they saw that

it was almost the same as one of the chains of haemoglobin.

And myoglobin from a sperm whale-- the haemoglobin here

was from a horse--

but, it showed an evolutionary connection.

And this was the first time that people had seen evolution

working at a structural level, at a molecular level.

And this is a type of insight that we get from structural

biology all the time now.

Previously, it had really just been looking at morphological

similarities in the external appearance of different

animals, and different species of plants, and so on.

Now, we can look at and study evolution at

the molecular level.

So, this is another benefit of the technique of


So, a later structure, lysozyme, was solved here, in

the Royal Institution, in a group

involving David Phillips.

And again, this was now the time when

Lawrence Bragg was director.

And this sketch of the protein molecule is actually done by

Lawrence's own hand--

he was quite a talented artist.

And what a pleasure it must have been for him--

and this is 1965, thereabouts--

50 years after he and his father had first, sort of,

worked on this technique and solved the structure of just

two atoms, here was the structure of lysozyme.

This is an enzyme.

This one is from chicken eggs-- it actually helps the

chicken egg to fight off bacteria because it chews up

the bacterial cell wall.

This is a modern representation of it, and we

can now see, in atomic detail--

the orange molecule is a small segment that looks like a

bacterial cell wall-- and so, we can see how the enzyme

works-- how it catalyses pulling apart this molecule.

So, crystallography gives us incredible molecular,

atomic-level insights into the workings of biology.

And it has just gone from strength to strength.

That result was from the 1960s.

These days, we have much better kit, but we're

basically doing the same thing.

We are still growing crystals, we are still collecting x-ray

diffraction patterns, and we are still solving structures.

The kit's a bit better.

So rather than a Crookes tube, which is what the Bragg's

started out with, we've got one of these.

This is a particle accelerator.

It nestles in the Oxfordshire countryside near Didcot.

This is the diamond light source.

It is the most expensive scientific facility that

Britain has built in the last 10 or 20 years, and it is

something that we should be very proud of.

It's an excellent, world-class facility.

So now, we just grow our crystals, and we take them to

diamond, and we fire x-rays at them.

And instead of the, sort of, early patterns, like

haemoglobin, this is now a

modern-day diffraction pattern.

And as we illuminate the crystal, we can rotate it.

And instead of collecting data on photographs, we now have

electronic detectors that capture the diffraction almost

in real time.

And, from all of these different images, we can then

work out structure after structure.

And so, the technique really has matured,

really has grown up.

From sodium chloride, which was just two atoms in 1912,

1913, to lysozyme--

which is probably a couple of thousand atoms-- in the '60s.

We now have structures like this.

This is actually a sodium potassium ATPase.

So, this is --.

This is salt.

And this is a molecule--

a transporter--

that sits in your cell membranes and regulates the

influx and efflux of the sodium and potassium ions to

maintain a healthy salt balance

within the living cell.

So, this is a gigantic structure.

However, it's not very big compared to this.

Now, this is the ribosome.

Now, the ribosome is a monster.

So, this is the ribosome from a yeast, which is

a eukaryotic cell.

It's a sophisticated type of cell, which has the same type

of organism that we are.

It's just we are a many-celled organism and yeast is only a

single-celled organism.

But, this is a truly gigantic structure.

Bragg's first structure had two atoms.

The ribosome has 404,714 atoms in this crystal structure.

It's truly a monster.

It might even be said to be akin to the monstrosity of

Hooke's flea.

I wish that it would capture the public imagination in

quite the same way, but it's worth bearing in mind that

this flea contains many millions of copies of a

molecule just like this.

The ribosome is an enormous and powerful machine, which

helps us to decode the genetic code and synthesises proteins.

So, from all of those techniques, from x-ray

crystallography applied to salt, we have moved to brand

new territory.

And so, we are learning all the time about life, and we

are learning about it in ways that are brand new.

And so, if we now just, finally, look at the ribosome

in the wild, so to speak, you can see where it takes its

place inside the cell.

And this is a beautiful painting from a

book by David Goodsell--

if I show it up here to give him the credit for it--

this shows the molecular interior of a cell.

This is bringing together all the knowledge we have about

what molecules are present and what structures they are

likely to have.

We haven't yet solved all these structures, but thanks

to crystallography, we very soon will be able to do so.

Finally, we are solving structures on Earth, but

crystallography has moved beyond this planet.

And on the Curiosity Rover that NASA sent to Mars last

year, there is an instrument that does x-ray diffraction.

And the robot arm can toss a sample of soil into this

device, and it can take an x-ray

diffraction pattern on Mars.

And so, x-ray crystallography has moved off-world, so to

speak, and is now telling us that the soil on Mars is a bit

like Hawaii.

I don't think the climate is quite as nice.

And this, again, is just a structure from this week's

Nature, published on Wednesday--

another monster.

This is a polymerase that actually makes the RNA that is

used to build the ribosome.

And so, we are making connection after connection at

the atomic level.

So, I hope you will now agree with me, that x-ray

crystallography is one of the most powerful methods that

science has produced in the 20th century, and has truly

allowed us to see the world in a completely different light.


The Description of Seeing Things in a Different Light: How X-ray crystallography revealed the structure of everything