1. 1 Convection in high res probes Pete Gierth, Bruker UK Worldwide application meeting 2015

2. Convection Wikipedia says: Convection is the concerted, collective movement of groups or aggregates of molecules within fluids (e.g., liquids, gases) and rheids, either through advection or through diffusion or as a combination of both of them. We normally mean simply advection, and treat diffusion separately Source is density gradient, arising from temperature gradient Extent is affected by viscosity amongst other things

3. Convection The main problems caused by convection: Distortions of diffusion measurements Losses in gradient coherence selected experiments Distortion of gradient echo profiles in gradient shimming

4. D=4*10 -10 m 2 /s 0.05 mm/s 0 mm/s 0.1 mm/s 0.2 mm/s 0.5 mm/s Effect of convection on diffusion measurements

6. Rayleigh–Bénard Convection http://commons.wikimedia.org/wiki/File:ConvectionCells.svg

7. Onset of convection - Rayleigh Number Case of infinitely long cylinder Acceleration due to gravity Thermal expansion coefficient Density Specific heat capacity Tube radius Viscosity Thermal conductivity

8. Onset of convection – critical temperature gradient Case of infinitely long cylinder dT/dZ (CHCL 3 ) ~ 0.1-0.3K/cm dT/dZ (DMSO) ~ 0.6-1.8K/cm dT/dZ (H 2 O) ~ 2.4-7.2K/cm

9. Horizontal temperature gradients Hadley convection

10. Sample:water for convection test CH 2 Br 2 in nematic liquid crystal for T measurement 5mm tube 10mm tube 10mm tube contains air or Krytox oil

11. 1H1H GzGz t1/2 Evolution of D HH only

12. ~0.3 K ~0.4 K 2cm? air oil dT/dz = 0.03K/cm!

13. 14mm filling

15. 450l/hr 200l/hr 1.5W dec. 1.5W 6W 320K 5mm/s 3.3mm/s

16. Temperature profile in the X-Y plane Nominal temperature 320K Strong gradient along Y axis Temperature range +/- 0.14 = 0.7K/cm Theoretical v max = 2.1mm/s in CDCl 3

18. Horizontal temperature variation ~1K/cm

19. Data on Bruker probes Fitting V=0.06mm/s V=0.2mm/s V=1.2mm/s

20. Data on various Bruker probes Chloroform sample Older TXI, dual flow, 400l/hr Older SEI, dual flow, 400l/hr No chiller

21. Data on various Bruker probes Chloroform TXI, made 2012? BCU-I 535 l/hr Shim temp ~306k

22. Data on various Bruker probes Chloroform BBFO (not Smart) 535 l/hr

23. Data on various Bruker probes Chloroform BBFO (not Smart) 400 l/hr T gradient <0.2K /cm @ 323K

24. Data on various Bruker probes Chloroform BBFO (not Smart) No VT gas at all 0.012mm/s 0.02mm/s

25. Data on various Bruker probes Chloroform SmartProbe, 3mm tube 400 l/hr T gradient about 0.1K/cm @ 323K as measured in 5mm tube V max <<0.1mm/s around room temp V max = 0.1mm/s @323K

26. Data on various Bruker probes Chloroform SmartProbe 500 l/hr, no chiller (400 MHz), 535l/hr, BCU-I (500MHz) 5mm tube

27. Data on various Bruker probes Chloroform Prodigy 298K, 535l/hr Vmax = 0.24mm/s

28. “Cryoprobe” simulated by imposing artificial T gradient 200 l/hr gasflow Slice selective experiment on MeOD Each row refereced using CD 2 H peak Gradient of OH peak position shows T gradient Z shim correction on water sample = 24 units

29. Chloroform 2.2mm/s D 2 O 0.14mm/s “Cryoprobe” 0.3K/cm temperature gradient Chloroform – 3mm tube 0.08mm/s

30. Conclusions Convection is indeed ubiquitous Extent is rather probe dependent Under most conditions significant enough to affect measured D values Use 3mm or sapphire tubes if necessary If in doubt, measure!

31. Acknowledgements Aitor Moreno Jerome Coutant

32. www.bruker.com