MIT Reports to the President 1996-97


During the past year, the Francis Bitter Magnet Laboratory (FBML) has made notable advances in several areas of science and engineering involving high magnetic fields. The research program in Magnetic Resonance (primarily nuclear magnetic resonance (NMR), but also including electron paramagnetic resonance (EPR)) has continued to grow and is now the largest effort at the FBML. The program continues to be funded primarily by the NIH and DOE, and involves ~20 NMR and EPR magnets and spectrometers some of which have been custom designed, some acquired commercially. These include a wide bore 360 MHz, two wide bore 500 MHz, three 600 MHz, and two 750 MHz NMR systems. In addition, we now operate the sole 140 GHz EPR spectrometer in North America, and have recently acquired a new widebore magnet with an expanded range sweep coil for this system.

A web site describing some of our research can be found at the following URL:



We used our newly developed technique of high resolution NMR scattering to make the first direct measurement of rate of spin diffusion through a homogeneous dipolarly coupled solid (single crystal calcium fluoride). This measurement has been the focus of theoretical studies for the past 30 years and a variety of failed experimental studies. In many respects it is an ideal test system for multi-body theories and our experimental results should help reinvigorate this area.

We used our new approach to high resolution NMR of semi-solids (based on a combination of magic angle sample spinning and magnetic field gradients) to quantify the correlation of molecular diffusion, compartmentalization and variations in the local magnetic susceptibility. These methods will find use in characterizing the average local structure of biomedical systems at sub-micron length scales.

We have extended the recently developed paradigm of "NMR computation" to demonstrate that quantum computing is accessible through the complexities of an ensemble and that fundamental quantum behavior is still observable following the ensemble average.


High Resolution NMR Microscopy - We have continued to push for higher resolution and sensitivity NMR microscopic images, particularly through the implementation of diffusion insensitive slice selection methods, rapid constant time imaging schemes based on reduced k-space sampling and the characterization of micron structures via local gradients.

RF Gradient Spectroscopy - We continue to explore the applications of novel radio frequency gradient hardware and methods to improving structural and dynamic studies of bio-molecules by NMR. Recently we have explored the applications of RF gradients to bi-linear rotation sequences (BIRD and TANGO) and have developed a more robust two channel NMR probe.

High Resolution NMR Scattering - We have completed the spin diffusion measurements discussed above, and an improved version of the high gradient strength probe. We are now turning our attention to measures of mesoscopic spin dynamics.

MRI of Soil Remediation - (with Prof. P. Culligan-Hensley, Civil and Env. E). We have completed a set of measurements that show the images of oil displacement by a water stream in a packed bead system. This provides the only direct, non-invasive, three-dimensional data on the efficiency of oil removal and is needed as a link to microscopic models of the system.

Applications of NMR to Liposarcoma Grading - (with Dr. S. Singer, BWI and Dana Farber) We have seen that gradient, HR-MAS is indeed a powerful means of exploring the chemistry and compartmentalization of sarcomas and have expanded the study to include cultured cell lines.


Dynamic nuclear polarization (DNP) provides a mechanism for transferring the high spin polarization of unpaired electrons to relatively less-polarized nuclear spins. Sensitivity enhancements of up to 3 orders-of-magnitude are possible with this technique. We have recently achieved reproducible enhancements of 50 for CP/MAS (cross-polarization/magic angle spinning) spectra of the 18.7kD protein T4 lysozyme in frozen solution with the spin label TEMPO as the source of unpaired electrons. Polarization is first transferred from electron to proton spins under microwave irradiation; proton spin diffusion and subsequent CP to low-gamma spins (e.g., 13C, 15N) allows uniform polarization across the sample. The electron-nuclear polarization transfer is primarily driven by the thermal mixing effect, where off-resonance irradiation of the TEMPO EPR (electron paramagnetic resonance) line perturbs the electron dipolar bath, and electron-electron-nuclear spin flips drive polarization transfer to the nuclei. The efficiency of this process depends critically on the homogeneous nature of the EPR lineshape. Spectral spin diffusion can render the typically inhomogeneous EPR line effectively homogeneous for the purposes of this effect. The timescale and spectral variation of the spin diffusion effect was studied via ELDOR (electron-electron double resonance) and electron spin-echo relaxation measurements. Experiments varying radical concentration, magnetic field strength, and temperature reveal additional features of the spin diffusion effect and the polarization transfer mechanism.


We extended our understanding of the vector pictures of MAS (magic angle spinning) to cases involving an axially asymmetric tensor. We also developed an extended treatment of frequency selective heteronuclear recoupling sequences and MAS which is useful for distance measurements between 13C and 15N and 31P and 13C. We initiated experiments to obtain high resolution spectra of I>=3/2 nuclei via multiple quantum excitation. The initial results employed two pulse sequences, and we have recently improved the methodology via a 3Q to 1Q transfer scheme referred to as "RIACT" (Rotationally Induced Adiabatic Coherence Transfer). The RIACT approach should be especially useful for 17O NMR in biological systems. We reported the initial MAS spectra of H217O. Resolution is always a problem in NMR spectra and we explored the possibility of significantly increasing the resolution via 3D 15N-13C-13C chemical shift correlation spectra. The spectra exhibit some unique features (very asymmetric 2D slices) and a theory was presented explaining these features. Efforts are also underway to determine peptide torsion angles via correlation of 1H-15N and 13C-1H dipolar tensors. The advantage of this approach over distance measurements lies in the fact that distances are often not sensitive functions of [phi] and [psi]. Direct measurements of torsion angles circumvents this problem. We published findings describing rotational resonance tickling (R2T), a new approach to measuring distance, which employs selective excitation and is capable of probing a long distance (weak dipole coupling) in the presence of a short distance (strong coupling).

We have made significant progress in developing methods that will be useful for determining the structure of large and/or insoluble proteins or peptides, we have improved the resolution with new 3D sequences. We have continued experiments on I>=3/2 nuclei via multiple quantum excitation, which could be widely applicable to biological systems. Finally, we have described two new approaches to structural studies: torsion angle measurements and selective distance measurement.


We published findings describing the origin of the line broadening in membrane peptide and protein spectra. While the experiments involve a small molecule, the important point is that it exhibits motion on the time scale of the 1H decoupling. In this case the decoupling interferes with the recoupling. We have developed a new approach to perform heteronuclear recoupling. We analyzed 2H lineshapes for a flipping H2O molecule which will be useful for describing membrane protein dynamics. We developed a new approach to dipolar recoupling in lipid bilayers used to measure segmental order parameters. We investigated the structure of the Schiff base counterion interaction with an emphasis on the 13-cis form and the L-intermediate via low temperature experiments. We continued our study of the effect of diffusion of peptides on spectra in lipid bilayers - in this case Gramicidin-A. We also published new approaches to measuring torsion angles via multiple quantum NMR effects and an approach to 3D spectroscopy of membrane proteins.

Our results demonstrate convincingly that we are able to measure distances in membrane peptides and proteins and establish the mechanism of the intensity losses in the spectra. The results establish the structure of the Schiff base in the L-intermediate of bacteriorhodopsin.

The condensed matter physics effort at the FBML is focused on spin-dependent tunneling between magnetic films of CrO2. The films have potential applications in the digital electronics industry, automobiles and medical diagnostics. Last year, we reported that the Stark-Faraday effect in GaAs-AlGaAs structures at 0.8um had been observed. The effect has possible optoelectronic device applications and a patent was recently granted for devices based on the effect.


This program's principal objectives have been to advance the theoretical understanding of low-temperature sliding behavior and to expand the cryogenic tribology data base. The program has been highly successful over the years and provided a database that is unique and very useful for the design of superconducting magnets and cryogenic devices. As research activities in the past few years have shifted, particularly in the US, from design and operational issues of low-temperature superconducting (LTS) magnets to those of high-temperature superconducting (HTS) magnets, mechanical disturbance---friction heating---that is of paramount importance for LTS magnets and has been the basis for our research activities, has become less pressing. Because of this shift in research emphasis from LTS to HTS magnets, while still continuing to generate additional cryotribology data, we have initiated a new HTS research program under this project: a new magnetic levitation system we call "electromaglev," in which an HTS bulk sample, e.g., YBCO, is levitated stably in a DC magnetic field generated by electromagnets placed underneath the floating object. During the past year a comprehensive study, both theoretical and experimental, has been completed and results will appear in Cryogenics.


We completed, and published results of, an experimental and analytical study of two-dimensional normal zone propagation in pancake test coils, wound with silver-sheathed BSCCO-2223 tapes. Two test coils were studied in detail, one having 3 layers and the other 8 layers. Each test coil was housed in an adiabatic environment whose temperature (20-70 K) was controlled and maintained by a two-stage G-M cryocooler and placed in a background field (0-6 T) generated by a Bitter magnet. With a test coil carrying a transport current (0-200 A), a local heat disturbance was applied by a heater attached to the coil's outermost layer. The coil's resulting electrical and thermal responses were recorded with voltage taps and thermometers attached to the coil. A normal zone propagation code was developed to simulate each coil's voltage and temperature responses for both quenching and recovering events. The code solves the nonlinear transient heat diffusion equation in two-dimensional cylindrical coordinates with a finite difference method. As an application of this code, a two-coil system, each coil comprised of one double pancake wound with silver-sheathed BSCCO tape, was studied for its quench behavior as one of the coils was driven normal locally. The simulation results indicated that the value of a shunt resistor connected across the terminals of each coil had a profound effect on the level of hot-spot temperature reached in the quench initiation spot. The research is continuing with layered HTS coils to study normal zone propagation in 3 directions. Intermagnetics General Corporation (IGC) and Sumitomo Electric Industries have generously donated the BSCCO-2223 magnets used in our experiment.


A US patent has recently been granted to one of the MTD members on the concept of a "permanent" HTS magnet system. The newly patented system combines the simplicity and ease of operation of a ferromagnetic permanent magnet with the strength capability and versatility of an electromagnet through the use of high-temperature superconductors and a new approach to the operation of superconducting magnets.

Once energized and producing a desired field, the system, without being coupled to a cryocooler for refrigeration, can maintain the field for a long period. This cryocoolerless lightweight, "permanent" HTS magnet is particularly suitable for on-board and/or portable applications, where "permanence" in most applications means a duration of hours, days, weeks, months, or perhaps years. Note also that unlike a conventional permanent magnet, this "permanent magnet" can also be used as an energy source. Two other innovative ideas incorporated into the newly patented system are "recooling" and "recharging" capabilities. That is, the system is designed to be recooled while it maintains its constant field so as to make, through periodic recooling cycles, the field literally permanent; it may also be recharged if its upper operating temperature is exceeded and the field decays.

Condensed matter physics research in the areas of thin film magnetism, semiconductors and superconductors is being conducted with benefit to fundamental knowledge as well as future application. For example, our recent success in spin dependent tunneling, anticipated for over 25 years, has not only opened fundamental questions but also is extremely promising for ultra high density recording (>10Gbits/in2), non volatile memory elements, and sensors alike. Several major companies including IBM, Philips, Motorola are working on this new technology. For far future atomically resolved storage (>1Tbit/in2), semiconducting phase change materials are being explored, with funding from a major company. Academically, in addition to graduate students, post docs, undergraduates, even high school students participate in the research program. National and international collaborators from universities, national labs and industries are also involved. The research has been cited in various magazines and newspapers as well as scientific journals.

The Bleomycins (BLMs) are anti-tumor antibiotics which are presently used clinically in the treatment of head and neck cancer and testicular cancer. There, cytoxicity is thought to be related to their ability to bind to double-stranded (ds) DNA and cause ds-lesions which are difficult to repair. While the chemistry of the DNA-cleavage reactions have been elucidated in some detail the basis for molecular recognition of a dGpPy sequence is still unknown. We have embarked on a program using solution-state two-dimensional NMR methods to determine the structure bound to an oligomeric DNA containing a single recognition site. We have chosen to examine a cobalt hydroperoxide analog of BLM which is an analog of the ferric peroxide BLM thought to be the activated species in-vivo. This species is chemically stable in the dark, it cleaves BLM with the same sequence selectively as iron BLM, its ligands are exchange inert, and it is diamagnetic. We have carefully chosen two oligomeric DNAs for examination, each containing a single BLM binding site and cleavage site. We have recently determined the first sequence-selective Kds for CoBLM which range between 5x10[macron]6M and 10[macron]7M. The oligomers being examined are d(CCAGGCCTGG)2 (1) and d(CCAGTACTGG)2 (2). A titration of CoBLM with each of these oligomers reveals a 1:1 complex in slow exchange on the NMR timescale. We have recently published the structure of CoBLM A2 green complexed to (1), above (J. Am. Chem. Soc. 118, 1268-1280 (1996)). We have also recently solved the structure of CoBLM A2 brown form. In this case, the axial ligand is H20. Thus far we have been unable to define the screw sense of this analog. We have also acquired the data to solve the structure of phleomycin, an analog of BLM in which one of the thiazolium rings is reduced. Solutions of these structures should explain their sequence specificitivity in bonding to DNA.

We have been applying the spin-polarized tunneling technique to the study of unusual spin states in superconductors. These states include non-equilibrium conditions caused by the injection of spin-polarized carriers into a thin film superconductor and the Fulde-Ferrell state predicted to exist in special cases for superconductors at low temperature and high magnetic field. As part of this program, we have been investigating the tunneling properties of the conducting ferromagnet chromium dioxide for possible use as a source of spin-polarized electrons. Band structure calculations imply that this material might be a half-metallic ferromagnet and thus would be a superior electrode material for spin-polarized tunnel junctions. This work is supported by the NSF. Personnel include Paul Tedrow of the Magnet Laboratory staff and Professor Katsuhiko Suzuki of Miyagi National College of Technology, who is supported by the Ministry of Education of Japan.

We have published a preliminary model of the HIV TAR RNA in complex with argininamide. However, several features of this RNA are poorly resolved. The wild-type TAR RNA has a three nucleotide bulge (UCU) that is critical for formation of the correct structure such that the HIV tat protein can bind. A naturally occurring TAR variant has a two nucleotide (UU) bulge, and we have prepared this RNA for structural characterization. It binds to argininamide somewhat more tightly than the UCU bulge TAR, and we observe several new inter- and intramolecular NOEs in the bulge region. After collecting a huge number of 2D-, 3D-, and 4D- data sets on 13C-labeled RNA, and 15N-labeled RNA, we have obtained a large number of distance constraints. There are over 700 NOEs that define the structure of this 30 nucleotide RNA, including 20 intermolecular NOEs to the argininamide ligand. We have completed the structure determination, and the paper describing the structure is in press. The first U nucleotide in the bulge region forms a base triple with an A-U base pair in the upper stem, and the arginine binds immediately below this U base, interacting with a G-residue in the major groove. The guanidinium group is stacked between two bases, forming an arginine sandwich.


The Magnet Technology group completed the winding of the major part of the 45T hybrid magnet to be installed at the National High Magnetic Field Laboratory in Tallahassee. In addition, the past year has seen the acquisition of five new magnets for the magnetic resonance research effort. These include a 104 mm, 500 MHz magnet for solid state spectroscopy, a 52 mm, 600 MHz magnet and spectrometer for microscopy and a similar system for solution NMR, and a 62 mm, 750 MHz magnet for solution and solid state experiments Finally, we have acquired a 125 mm 5T magnet with a sweep coil for extended high field EPR experiments.

A new, one-of-a-kind Molecular Beam Epitaxy (MBE) system was built to explore the frontiers of spin tunneling, surfaces and interfaces of magnetic and semiconducting materials on an atomic level. This system will enable us to improve our fundamental understanding as well as future digital storage and sensor view point.

Plans are underway to consolidate FBML resources into one building as facilities currently housed in building NW17 (including two 750 MHz NMR magnets, a 200 MHz/40 cm NMR magnet, a wet lab and electronics shop) will be moved to building NW14. This consolidation will do much to enhance the FBML research center and to make space available for the LIGO project to move into when building 20 is replaced.


The Laboratory contributes to undergraduate education by participation in the Undergraduate Research Opportunities Program (UROP) a program that encourages and supports research-based intellectual collaborations of MIT undergraduates with Institute faculty and research staff. In addition, the laboratory has 35 full time graduate and 14 postdoctoral students who are performing research.

Dr. Susan S. Pochapsky was hired as a Sponsored Research Staff member to improve the resources of the magnetic resonance facility.


Plans are continuing for a major expansion of MRI activities involving the Harvard/MIT Division of Health Sciences and Technology (HST). HST is in the process of recruiting senior faculty members in the area of functional MRI, and the FBML Director is working closely with HST faculty toward the recruitment of renowned individuals in this field.

A proposal to the NIH for the development of very high field, wide bore NMR magnets is awaiting a decision. If funded, the Technology Group would pursue the program for three years.

Preparation of the competitive renewal to fund the Center for Magnetic Resonance is underway and will be submitted to the NIH before the end of 1997.

Robert G. Griffin

MIT Reports to the President 1996-97