In most of the textbooks on NMR, we come across the fact that under the 1.4 Tesla external magnetic field, the precession frequency of 1H (proton) is 60 MHz. Can anyone tell why it's always mentioned as 1.4 T instead of 1 T?
1.4 Tesla is the maximum field strength achievable in a conventional configuration (natural magnet-at the time) which was known in the early stage of NMR development (before the manufacture of superconducting magnets).
The resonance frequency of 1H (in a magnetic field) is a property of the single proton (or - more correctly- of the quarks making it up). It is experimentally determined (as the theoretical understanding of the atomic nucleus is limited).
When I was lecturing this kind of stuff I used 2.349 T which gives a resonance frequency of 100.0 MHz (for TMS), but for 13C the resonance frequency in 2.349 T is 25.15 MHz (all NMR active nuclei will have different frequencies).
This strenght of magnetic field was achievable in the first NMR systems in the 50s and 60s of the 20th century with permanent iron magnets. Undoubtedly they were produced bad spectra, with poor resolution (often spectra with so called ABC systems), but this was the level of NMR equipment in 1960. Don't forget, that NMR spectroscopy was discovered in 1946 from Purcell and Bloch.
Permanent magnets did not appear for NMR until 1967 (Varian T-60).
The early commercial spectrometers used electromagnets. At the time, a 1.4 T (60 MHZ) electromagnet magnet was near the practical electromagnet field-strength limit and divisible by 10 (a marketing decision).
.
The limits on maximum practical field strength were determined by two factors. The magnet frame had to be substantial in order to achieve and sustain field homogeneity. The magnet pole-face distortion had to be low enough to accommodate shim coil size in the narrow probes required to minimize pole-pole distance. Other factors were electricity consumption and limiting magnet water cooling equipment costs.
Later on the instrument companies realized 2.3 T (100 MHz) electromagnets were a practical limit due to pole-face distortion. Even if the pole-face distortion problem could be solved higher fields also required absurd electromagnet size and weight, unacceptable power consumption and extensive magnet cooling equipment.
Eventually Varian developed the EM-390; a 2.1 T (90MHz) CW spectrometer with a permanent magne that was small enough to be practical. Even then there were problems with the magnet weight. Early EM-390 installations were problematic because shippers and laboratories did not have equipment that could properly handle the magnet weight. Many initial deliveries had to go back to the factory because the pole faces were hopelessly misaligned by rough handling. The problem was solved by informing customers about handling requirements and installing accelerometers on magnet shipping crates.
a "B"-field of 1 T would produce a resonance frequency "w" for a "g" given for 1H of 42.6 MHz and one may ask: what is special about the 42.6 MHz?
So from physics point of view nothing special with 1.4 T (60 MHz) just "numbers" related by w=g*B and what counts was said above (availability and feasibility, but also investment needed for instrumentation)....