No abstract

No abstract

NMR-microprobes based on solenoids and Helmholtz coils have been microfabricated and NMR-spectra of mammalian cells have successfully been taken. The microfabrication technology developed for these probes consists of three electroplated copper levels for low resistance coils and three SU-8 layers for the integration of microchannels. This technology allows fabricating solenoids, Helmholtz and planar coils on the same wafer. The coils have inner diameters in the range of 160 to 400 microm and detection volumes of 5 to 22 nL. The solenoid and Helmholtz coils show improved RF-field characteristics compared to a planar coil fabricated with the same process. The fabricated solenoid has a particularly low resistance of only 0.46 Omega at 300 MHz. Moreover, it is very sensitive and has a very uniform RF-field, but shows large line width. The Helmholtz coils are slightly less sensitive, but display a far narrower line width, and are therefore a good compromise. With a Helmholtz coil, a SNR of 620 has been measured after one scan on 9 nL pure water. An NMR-microprobe based on a Helmholtz coil has also been used to take spectra of CHO cells that have been concentrated in the sensitive region of the coil with a mechanical filter integrated into the channel.

Planar microcoils with diameter ranging from 20 to 1000 μm I.D. (130-1130 μm O.D.) are evaluated for their applications in NMR spectroscopy. The coils are first overfilled with a standard sucrose solution and compared against each other. Coils with smaller I.D. (≤100 μm) perform extremely well. One hypothesis is that as the coils get smaller the volume occupied by the copper turns increases relative to the open I.D.; as such a large proportion of the sample is brought in close proximity to the coil turns and likely gives rise to strong sample-coil magnetic coupling, which increases the signal. The applications of the planar microcoils are demonstrated on Cypselurus poecilopterus (fish) and Daphnia magna (water flea) eggs. A single D. magna egg on a 50 μm coil yielded at least 3000 times the mass sensitivity (∼9,000,000 time saving) when compared to a 5 mm probe. This value could be at least 4 times higher if the B homogeneity of the coils could be improved. With the current design, 80% of the signal is lost in multiple pulse experiments that rely on phase inversion and signal cancellation between scans. The data were extrapolated to predict that biological samples as small as ∼4 μm may become accessible via planar microcoil designs. To fulfill their potential for in situ metabolic screening, specialized magnetic susceptibility matched sample holders that restrict the sample to the homogeneous B field region (i.e. within the 90% RF field) of the coil and advanced experiments that narrow spectral lines, suppress lipids and disperse signals into multiple dimensions will be required.

Nuclear magnetic resonance microscopy at an isotropic resolution of 3.0 lm was realized by using dedicated hardware such as RF surface microcoils, a planar triple-axis gradient with 6,500 G/cm, and a static magnetic field of 18.8 T. Purely phaseencoded constant time imaging was used to allow increasing the gradient strength for the suppression of diffusion effects without reducing the signal-to-noise ratio. For this method the relationship between gradient strength and true spatial resolution was investigated, and an empirical formula is provided that is useful for practical applications. The characteristics of the different hardware components were investigated experimentally. Furthermore, microscopic phantom images were acquired and evaluated for their true resolution. It is demonstrated that the use of sufficiently large gradients enables suppressing diffusion-related loss of spatial resolution.