If you are 40 meters away from a signal with a wavelength of 6000 km, you are definitely within the near field of the radiator. This means that the electric and magnetic fields can be (and will be) vastly different. The magnetic field will reflect the current in the wires, and the electric field, the voltage. Both of them will be problematic for detecting zero crossings.
The magnetic field is problematic for detecting zero crossings because in an electric power system, current and voltage do not need to be in phase, and commonly are not. Furthermore, current does not need to be sinusoidal, even when voltage is sinusoidal. If, however, you wish to monitor the magnetic field for some reason or another, a coil of wire will serve that purpose.
The electric field will be problematic for detecting zero crossings because of the means you will require to pick up the electric field. To pick up electric field, you should use two large conductive plates. The ground could theoretically be used for one of the plates, but I would be wary of what might be happening to fields in the ground, so let's assume two metal plates. The plates could theoretically be aluminum foil, but any movement in the plates will introduce noise into your system. So, the aluminum foil would need to be supported somehow to make the plates stiff. Perhaps a thin sheet of wood would work. So, your receiver will in essence work on the capacitance between your plates and the voltage source. Alas, a capacitor and the impedance of the voltage source together form a high pass filter. Even slight deviations from a sinusoid in the source will be "magnified" in your received signal. The following oscillograph shows the results of an experiment I did picking up the electric field from a extension cord (with no load) fed by an apparently clean sinusoidal mains voltage.
As you can see, it is pretty noisy. You could compensate by adding a low pass filter. This would remove some of the high frequency components of the signal. However, a problem will still remain. The intrinsic high pass filter, and the additional low pass filter will almost certainly introduce phase shift in your signal. How you will ensure the net phase shift is negligible is a problem that you will need to resolve if you wish to recover the zero crossing points of the signal.