Automating the single crystal x ray diffraction experiment

(1)

Automating the Single Crystal X-Ray Diffraction

Experiment

Mark E Light

School of Chemistry

University of Southampton, UK

(2)

Presentation Outline

Reasons for automation

Requirements for an automated system

Main challenges in building such a system

Work flow to clearly define the problem

Building and linking of system components

Inline evaluation

Software development

Output of results

Ongoing and future developments

Animations

Acknowledgments

(3)

Why Automate?

General reasons for building an automated system:

Maximise instrument usage – can do lots of shorter collections

overnight

Prioritise crystallographers time – let the robot do the tedious work

Utilise the ‘data fit for purpose’ ideas

Allows easy screening of many samples

Ideal for polymorphism studies

Allows for remote access

(4)

Requirements for an Automated System

Once started should not need human intervention

Ability to reliably place the crystal in the beam

Obtain sensible unit cell

Reject poor samples

Automatically setup and perform data collection

Make sensible decisions about exposure time, frame angle etc

Allow configuration but with robust defaults

Run unattended in a safe manner

Clearly report progress and status

(5)

Main Challenges

1.

1. Automatic changing of the sample Automatic changing of the sample 2.

2. Hands free crystal centringHands free crystal centring 3.

3. Automation of the data collectionAutomation of the data collection 4.

4. Continual Inline evaluationContinual Inline evaluation 5.

5. Development of a graphical control interfaceDevelopment of a graphical control interface

These can be placed within the following work flow

(6)

Control Flow

Diffraction

Unit Cell

Success

Strategy

Data Collection

Data Process

System Y

PreScans

Yes

BruNo Mount

BruNo Unmount Setup via GUI

Sample Tray

No

(7)

1. Sample Changing

Mitsubishi Movemaster RV-1A 6 axis industrial robot with a CR1 controller – 30K

Internal control program accessed via serial port

(8)

2. Crystal Centring

Uses accurately constructed pips

Specially designed goniometer head employed

Largest collimator used to maximise beam size at crystal (0.6mm)

Some miscentring is allowable

Why not use motorised goniometer head and image recognition

software?

Considerable added cost

Added system complexity

Additional sample runtime

(9)

3. Automated Data Collection

As defined in the work flow – the following steps are carried out

Determination of crystal diffraction characteristics – using prescans -

evaluate evaluate

Determination of unit cell using phi/chi scans and DirAx -

Determination of unit cell using phi/chi scans and DirAx - evaluateevaluate

Carrying out of strategy calculation and data collection

Processing of data

Running of structure solution and refinement -

(10)

5. Control GUI

Requirements

-Data Collection:

Data Collection:

Well structured logic flow and failsafe error handling

Sensible automated parameter choices

GUI

Easy sample selection and meta data input

Highly configurable but with robust defaults

Clear progress status

Developed X-Tray

Written in python using tkinter and the Bruker-Nonius Collect

package.

(11)

X-Tray

(12)

X-Tray

(13)

X-Tray

(14)

Output of Results

For an automated experiment the ultimate aim is a completed structure!

Intermediate results are presented to the user:

html report containing full experiment details

hkl reflection file

Starting point shelx instructions file to present to user or System-Y

System-S best try at structure

System-Y results

Structure validation report

(15)

Ongoing and Future Developments

Improvements to the pip manufacture

Improvements to the goniometer head

Completion of the collision safety circuit

Inline ranking of samples – so that poor samples are evaluated and

then put at the back of the queue

Addition of a controlled environment for queued samples

Improved unit cell evaluation

Develop results dissemination