Because of the rapid rise of helium price in recent years, high-temperature superconducting (HTS) magnets are expected for helium free MRI systems. Two years ago, we started a project targeting to produce HTS ultra-high-field (UHF) MRI systems with REBCO (Rare-earth - Barium - Copper Oxide) tapes, known as second generation tapes. Although superconductivity characteristics of REBCO are higher than the other HTS tapes, a DC power supply necessary to maintain the field strength and strong shield current induced in the tapes can be causes of field instability and inhomogeneity. Therefore, as the first step, we developed a 1T-REBCO magnet with 200 mm room temperature bore and evaluated its magnetic field stability and homogeneity with dedicated field camera system with 16ch NMR probes.

The magnet was consist of two pairs of coils symmetrically arranged in Z-direction (Fig.1) and each coil of the inner/outer pair was composed of two/four single pancakes, respectively. The width and thickness of the REBCO tapes (Fujikura Ltd., Tokyo, Japan) are 5/1.3 mm, respectively and total length of the tapes was 3.2 km. The inner diameters of all coils were 400 mm and the FOV region was a 100 mm sphere. Current of 285 A was necessary to achieve field strength of 1 T, and usually ramp-up duration was two hours.

The dedicated field camera [1,2] (Fig.2) was consists of 16ch NMR probes. All materials of the probes were carefully selected with considering their magnetic and electrical properties. A solenoid coil with a grass capillary (inner/outer diameters were 0.81/1.09 mm) containing water was put in a case produced by a 3D printer (Agilista-3100, Keyence Co., Osaka, Japan), and was connected to a resonance circuit attached on the case surface. Its resonance frequency was tuned to 42.58 MHz and it was connected to a dedicated spectrometer produced with OpenCORE NMR technology [3] via two 4ch-RF switches (ADG-904, Analog Device, Inc., MA, USA) which control paths of Tx/Rx signals between 16 probes and the spectrometer. All probes were arranged on a 120mm circular disk (fifteen of them were on a 100mm-circle and the rest at the center) with a rotating shaft to measure distribution of field strength at an azimuth angle of the FOV surface and the disk was rotated after each measurement.

Figure 3 shows field distributions measured three times on two days (the field was ramped up/down everyday). The pattern of the field is very similar but average strength differs every time, and the peak to peak homogeneities were 730 ~ 800 ppm. Sphere harmonic extension of the average field distribution in Fig.4 shows that Z¹ and Z² components were dominant but that Z⁵ and Z⁶ components were also contributed.

To evaluate the field stability, the disk was fixed at an azimuth and field strength changes at the sixteen points were measured at every 16 sec (1 sec for each probe) for 1.5 hours. The field changes in Fig. 5a shows that the field increased 2.3 ~ 2.9 ppm/hour monotonically. Instability remained in field changes after linear trend correction is supposed to originated from both instabilities of the field and probes. However, some pairs of probes, whose measured timings were successive, shows very high correlation (> 0.99) (Fig.5b) and so the corrected instabilities can be considered that the field instabilities were dominant rather than the probe one. Then, peak to peak instability of the field was considered to be within ±1 ppm.

Our simulation study (not shown here) indicates the Z¹ and Z² components of the field inhomogeneity seems to be originated from manufacturing error and the higher order ones does from the original design mainly. If these are true, this inhomogeneity problem would be solvable in future.

The monotonic increase of field strength differs slightly by positions probably because this is shield current origin and because changes of shield current differ by positions. Therefore, this increase would be predictable and correctable in future. The fact that the not-high correlation between data obtained two not-successive-probes indicates that field was changing in a few second, and this is also problematic. In future, these instability is expected to be suppressed well with huge inductance of UHF magnets,

Field stability and homogeneity of the middle size REBCO magnet was measured and, although they were not high enough for MRI magnets, of course, the causes investigated here are expected to be solvable in future.