Someone at a conference once asked if it was true that with remote detection you could do NMR from here to the moon. While this person may have only been joking, it is true that remote detection NMR enables the transportation of NMR information from one place to another, although we haven’t yet reached astronomical length scales. In our lab, remote detection has lead to an assortment of experiments where the study of flow through porous media or micro-channels is important. For instance, images of hyperpolarized ¹²⁹Xe gas flowing through sandstone rock[1] an aerogel[2] and a microfluidic chip[3,4] were obtained using remote detection. New hardware designed for ¹H imaging now permits the study of flow and mixing of fluids, such as water and ethyl alcohol, in a microfluidic chip.[5] Current work involves studying fluid mixing in microfluidic devices with various mixing geometries as well as studying diffusion of fluids across porous membranes.

Remote detection was developed because of constrains that NMR instrumentation introduces on the size and type of samples that can be analyzed and the sensitivity that can be obtained. In conventional NMR, the same coil is used to encode and detect signal, which requires a compromise of optimal parameters for each step. With remote detection these parameters can be chosen with unprecedented flexibility compared to a conventional NMR experiment. An NMR active sensor is used to survey a chemical environment and retains an NMR-encoded memory to be recovered at a later time and location. The encoding and detection steps are physically separated, thus distancing excitation and evolution from the acquisition of signal in both the time and space. This allows for separate optimization of parameters that influence the quality of the data, such as field strength, coil size and geometry, receiver bandwidth, and even the type of detector.[6]

Remote detection experiments work by using a sensor medium that can store information as magnetization. One typical setup is to place a porous sample in an NMR magnet equipped with field gradients and to flow the sensor medium (¹²⁹Xe, H₂O etc.) through the sample. As the sensor interacts with the porous sample, a series of RF-pulses and gradients can select a region of the sample. The sensor then travels to a more sensitive coil where the information can be recovered. A travel time curve shows how long it takes for the sensor to travel to the detection coil. A similar scheme can be employed by replacing the detection coil with another detector.

The detector can be a more sensitive Faraday coil such as a microcoil, a squid, a megnetometor, or an optical detector. Current work in our lab involves exploring and developing these alternative detection techniques.

Multidimensional applications of remotely detected MRI generally involve phase encoding multiple spatial and possibly other dimensions. Subsampling of such higher dimensionality acquisitions, in combination with sparse image reconstruction techniques, dramatically reduces the time required to acquire such images. We have demonstrated applications of subsampling and sparse image reconstruction in five dimensional velocimetric MRI experiments in microfluidic devices, and are also pursuing other approaches to accelerated multidimensional acquisition based on spatial multiplexing.

[1] Granwehr J, Harel H, Han S., Garcia S and Pines A., Phys. Rev. Lett. 95, 075503 1-4 (2005). (reviewed in Nature News and Views v.437, 488-489 (22 Sept 2005))
[2] Harel E., Granwehr J., Seeley J., Pines A., Nature Materials 5, 321-327 (2006).
[3] Hilty, C., McDonnell, E.E., Granwehr, J., Pierce, K.L., Han, S.I., Pines, A. PNAS 102, 14960-14963 (2005)
[4] McDonnell, E.E., Han S.I., Hilty C., Pierce K.L., Pines, A. Anal. Chem., 77, 8109-8114 (2005)
[5] Harel E., Hilty C., Koen K., McDonnell E., Pines A. Phys. Rev. Lett., 98, 017601 (2007)
[6] Moule A.J., Spence M.M., Han S., Seeley J.A., Pierce K. L. , Saxena S., Pines A., PNAS, U.S.A. 100 9122-9127 (2003)