The James Webb Telescope: it’s all done with mirrors

Click image for a larger  version. The James Webb Telescope has reached the L2 point!

Friends, Hurray! The James Webb Space Telescope (JWST) has deployed all its moveable parts and reached its lonely station at the L2 point, far beyond the Moon.

In a previous article I mentioned that back in 2018 I had been fortunate enough to meet with Jon Arenberg from Northrop Grumman, and to see the satellite in its clean room at their facility in Redondo Beach, California.

  • In that article I outlined in broad terms why the satellite is the shape it is.
  • In this article I want to mention two other people who have made key contributions to the JWST.

I was fortunate enough to meet these people during my ‘career’ at NPL. And as I hope to explain, they have taken manufacturing and metrology to the very limits of what is possible in order to make a unique component for the JWST.

It’s all done with Mirrors

The 18 hexagonal mirrors of the JWST are iconic, but in fact there are many more mirrors inside the telescope.

JWST uses mirrors rather than lenses to guide the light it has captured, because at the infrared wavelengths for which the JWST is designed, glass and almost all other materials strongly absorb i.e. they are opaque!

In contrast, during reflection from a metal surface, light only enters the material of the mirror in a very thin layer at its surface.

Consequently, mirror surfaces can guide light of any wavelength with very low absorption.

Form and Smoothness

The creation of a mirror surface requires a machining operation in which a metal component – most commonly made from aluminium – is cut into a specific form with an exceptionally smooth surface.

  • The surface form must be close to the shape it was designed to be.
    • Otherwise the light will not be directed to a focus and the images will be blurred.
  • The surface roughness – the ‘ups and downs’ of the surface – must be much less than the wavelength of the light the mirror must reflect.
    • Otherwise the light will be ‘scattered’ from the surface and very dim objects will be obscured by light scattered from nearby bright objects

The large mirror surfaces on the primary and secondary mirrors are manufactured in a complex process that involves machining the surface of the highly-toxic beryllium metal, and then painstakingly grinding and polishing the surface into shape. Once completed, the finished surface is coated in gold.

Each step in the manufacturing process is interspersed with a measurement procedure to assess the roughness of the surface and the closeness to its ideal form. Ultimately the limit to achievable manufacturing perfection is simply the limit of our ability to make measurements of its surface.

For small mirrors, this polishing process is typically not possible and the surface must be cut directly on a sophisticated lathe.

To achieve the mirror-smooth surfaces, the lathe uses an ultra-sharp diamond-tool which can remove just a few thousandths of a millimetre of material at a time and leave a surface with near-atomic smoothness.

But typically the required surface form is not part of a sphere or a cylinder. To cut such ‘aspheric’ surface forms on a lathe requires that the lathe tool move on a complicated trajectory during a single rotation of the workpiece: the solid geometry and mathematics to achieve this are hellish.

To achieve the required form, the trajectory that the tool must follow relative to the workpiece is calculated millisecond-by-millisecond in MATLAB, and then instructions are downloaded to the lathe.

The JWST is filled with mirrored surfaces of bewildering complexity, channelling infrared light via mirrors to measuring instruments. And slightly to my surprise – and perhaps to yours – I have been told that considering all the mirror surfaces inside JWST, more were made in the UK than in any other country.


One of the key instruments onboard the JWST is the Mid-Infrared  Instrument (MIRI) which analyses light with wavelengths from 5 thousandth of a millimetre out to 28 thousandths of a millimetre.

And inside MIRI, one of the most complex mirrored-components is a ‘slicer’ or ‘splitter’ component.

Its precise function is hard to describe: the figure below is from an almost incomprehensible paper. My opinion is that it is incomprehensible if you don’t already know how it works!

Click on image for a larger version. Figure 14 of “The European optical contribution to the James Webb Space Telescope”. See the end of the article for reference. On the right is ‘splitter’ which redirects light in an almost inconceivably complex pattern.

So let me have a go.

  • Imagine parallel light from the main telescope mirrors falling on to a square section of a parabola. It is a property of a parabola that this light will be directed towards a single point: the focus of the parabola.
  • Now imagine splitting the square into two, and preparing each half of the square as sections of two different parabolas with their foci in two different places. Now parallel light falling on the component will be directed towards two different locations – with 50% of the light proceeding to each focus.

Click on image for a larger version. Illustration of the function of the splitter component. The left-hand panel shows parallel light falling onto a fraction of a parabolic surface being directed towards a focus. The right-hand panel shows parallel light falling onto a fractions of two different parabolic surfaces, and being directed towards two different foci. The Cranfield splitter has slices of 21 different parabola – each within 10 nanometres of its ideal form!

Now imagine repeating this division and machining a component with sections of 21 different parabolas in thin slices each just a couple of millimetres across. This what Paul Morantz and colleagues manufactured at Cranfield University in 2012. There’s a photograph of the component below.

Click on image for a larger version. An early prototype of the splitter mirror the JWST in the display cabinet at Cranfield University. It is – by definition, almost impossible to a photograph a mirror surface. But notice that each of the 21 mirror surfaces reflects light from a different portion of the label in front of it.

Each of the 21 different surfaces had to conform to its specified form within ±10 nanometres (IIRC). And to verify this required measuring that surface with that uncertainty. Measuring a complex surface with this uncertainty is at the limit of what is possible: re-machining the surfaces to correct for detected form errors is just breathtaking!

At each of the 21 different different focus points is a separate instrument measuring at slightly different wavelengths.

The outcome of all this ingenuity is a single custom component weighing just a few grams that simplifies the optics of the instrument, allowing more weight and space to be devoted to measuring instruments.

Pride and Wonder

Back in 2013 I was honoured to work with Paul Morantz and his colleague Paul Shore on the creation of the Boltzmann hemispheres which were used to make the most accurate temperature measurements in history.

The two hemispheres they created were assembled to make a cavity with a precisely non-spherical shape with a form uncertainty below 0.001 mm at all points over the surface.

But before the Pauls could get to work on our project, they had to finish the splitter for JWST otherwise its anticipated launch date might be delayed. [As it happened, they probably could have taken a little more time ;-)]

After completing the splitter I remember the disappointed look on Paul Morantz’s face when I explained that the Boltzmann project ‘only’ needed form uncertainty of 0.001 mm.

I cannot imagine their pride at having constructed such a wondrous object that is being sent to this remote point in space to make measurements on the most distant, most ancient light in the Universe.

I feel proud just to have known them as colleagues.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: