"Shuttle Device for Use in a Shared Commercial NMR Instrument Version II " (December 2004)
by A. G. Redfield, Brandeis University

Return to Index


This is a detailed description of our device as modified by early 2004. The first part of this report was mostly completed in the summer of 2004 and the second part was completed Dec. 2004. Some late news is summarized at the end. Work on this device was started in 2000, and described initially in a series of reports posted on my Brandeis website within months of developments. A fairly complete description of Version I was written in early 2003 and was kindly accepted as a contribution to a special issue of Magnetic Resonance in Chemistry devoted to relaxation, by the Editor of that issue, Professor Josef Kowalewski [Magnetic Resonance in Chemistry 41:753-768]. Subsequently we published 3 research articles: (with M. Roberts, C. Qui, C. J. Turner, & D. Case) Biochemistry 43:3637-3650 (2004); (with M. F. Roberts) JACS 126:13765-13777, and PNAS 101:17066-17071 (both 2004). The first of these showed the general utility, using a DNA duplex octamer, of field cycling 31P relaxation to evaluate sizes of CSA vs dipolar interactions and correlation times; the last two are on membrane dynamics and structure by 31P relaxatuion. We have also presented 2D 15N T1 relaxation data on cyclophilin A, down to 3.5 Tesla, at meetings (with Elan Eisenmesser and Dorothee Kern, unpublished). The latter are in good agreement with standard theory.

The figures are in the separate PDF file that is also posted, so that you can print some of them for study, and to avoid having this text file slow to read and print.

These reports are published on the web and in print in the hope that they will facilitate use of this important method in other labs. I will be happy to help non-profit labs reproduce this device by answering questions, within limits. However, this is not a stand-alone report, it is a continuation of, and supplement to, my article in Magnetic Res. In Chem ("MRC"). I will assume you have thoroughly read my article, as well as this report, and if you have not I may lose patience when you ask many questions already answered in them. If you ask intelligent questions I will try to answer as best I can, and would like to know what mistakes I have made or what is really unclear.

As usual, Frank Mello, the Director of the Brandeis Machine shop, made many useful suggestions and carefully built these components.

This report is posted on the web because version II described below is more straightforward to make than version I, and because MRC did not have drawings and other information useful for actually building and using the device. It is not much faster or otherwise better than version I. Version III is now being planned and will be described in about a year if it works well, and otherwise will be forgotten.

Besides reading the MRC article fairly carefully, you should skim the present report a few times. I do not provide a table of contents, but you should realize that the report starts from the smallest most inner part (the sample) and works out to the shuttle tube, then up the stop tube and over to the things on the remote table. So look at Fig. 2 of MRC carefully and then, after skimming this report, go back and study the assembly diagrams. These are Fig. SH1, the shuttle and sample holder, and Figs ST1, Fig ST2, and Fig ST3. The drawings are copied from working drawings, but certainly have errors. You must understand these drawings in any case, especially if you have to modify the device for a Bruker or other spectrometer. Please tell me about errors. Our system uses a standard Oxford magnet installed by Bruker about 10 years ago and retrofitted by Varian with their room temp. shim coils. You use these drawings and information at your own risk. Please supply me with drawings of modifications needed for other systems such as Bruker, or any other useful information.

I apologize for use of English units almost everywhere, but conversion to metric would probably have introduced errors. Some readers will dislike the quality of the drawings , and anyone who makes better ones could send them to me. Most of these drawings have been redrawn by me from working drawings used by Frank Mello and I think they are easily understood by someone familiar with the conventions of mechanical drawing.

Many scientists will have trouble understanding this report and I strongly advise them to find a physicist or chemical physicist as well as a good machinist, to help them after they have initially skim this report and tried hard to understand it. Then they could ask their physical mentor to quickly go over these drawings and explain what I intended to convey. The same applies to understanding the very trivial electronics that I describe in the second part, aided by an electronics professional.

The version I system and parts of version II were largely financed by a grant from the NSF Chem. Instrument Program. Some of the present design improvements were financed by my current grant from the Petroleum Research Fund. Early essential support came from my former NIH grant, and from the Rosensteil Center at Brandeis. I contributed much free labor supported by TIAA CREF. Therefore I will be extra appreciative if researchers will kindly acknowledge use of these designs in papers that first utilize them in a useful way.

The earlier reports that I posted will soon be removed or compressed into a brief summary.

Breakage. I have not broken a probe during a run except through stupidity. We probably have used the 10 mm probe about 40 days (24 hrs.) or more, and the 5 mm probe 20 days. Recently we have had 2 or 3 five mm tubes break at the extreme lower end without destroying the 10 mm probe. (that is an advantage of using this probe). Possible reasons are: something was not right in a complicated version of the stop assemblies and the shuttle was irregularly hit on some edge; or it was because I was not using any lower susceptibility plug, just filling the lower end with excess sample that we had, to save the small amount of work to put in a lower plug. We will now always pour in this plug with epoxy, and I have enlarged the lower hole (part (3) on fig ST 4B. Through stupidity I broke a piece of plastic inside the 10 mm probe at the bottom without breaking the coil insert glasswear, costing about $2500 for Varian to repair. It was because I had the NMR tube much too low, and this is now nearly impossible. I broke the long glass shuttle tube once by not pulling the stop tube straight, when its end had a longer diameter machined into with tighter tolerance.

Of course I test off-line with no probe slid onto the bottom, and with both 5- and 8 mm tubes. I have not broken any tube during such tests (which are not lengthy), at the maximum pressure (about 0.2 bar vacuum, 0.5 bar higher pressure, 0.35 bar lower pressure with transit time nearly down to 100 msec transit time not counting bounces. I have had tubes break during non-tests (real runs) however, at the top where the tube enters the adapter, and in the middle also. By luck the probe did not break, perhaps because I then had a small hole, slightly smaller than the Varian 5 mm entry hole, at the bottom of the shuttle tube holder (see MRC). We stopped doing this because we have to remember to remove it when using 8 mm tubes in the 10 mm probe, but I should probably arrange to use it again. In one case I did not let the epoxy set long enough, that holds the tube to the adapter, and it failed. I now always wait overnight after applying epoxy.

Probes are as described in MRC.