3.3 Comparison of Rotatable Polarisation to Projection via Wire-Grid Polarisers
55
frequency-dependent elliptical pulse, which is approaching circular polarisation for
frequencies above 3.0 THz. This ellipticity occurs due to the finite transmission of the
component of the THz electric field parallel to the wires; this component is the first
time derivative of the incident field [19], and is therefore phase-shifted by a factor
of π/2 with respect to the orthogonal component perpendicular to the wires. The
orientation angle also deviates significantly from the intended angle of ψ = −35.0
◦ ,
by 2.0
◦ in the best case at low frequency, and with larger change as the transmitted
pulse becomes more elliptical.
As demonstrated in Fig. 3.6b, the WGP was then removed and the emitter rotated
to ψ in = −35.0
◦ , the same angle as that discovered to be perpendicular to the wires
of the WGP in the first section of this experiment. This simulates rotating the polarisation state by 45
◦ by rotating the emitter, allowing direct comparison of rotating
the polarisation state and projecting it to the same angle. The frequency dependence
of the ellipticity and orientation angle of THz pulses in this case is shown by the
red lines in Fig. 3.6c, d, respectively. The rotatable polarization method produces a
linear THz pulse over the whole experimental bandwidth when rotated to an arbitrary angle, with an ellipticity of ≤ 0.75
◦ and a statistical error (the precision) in the
ellipticity of < 0.05
◦ over the 0.3−1.5 THz range (< 0.1
◦ from 0.3 to 2.5 THz) after
20 repeated measurements. The accuracy of the orientation angle was defined as the
variation of ψ(ω) away from the mean [6], and was < 1.0
◦ between 0.3 and 2.5 THz.
The statistical error in the orientation angle was < 0.05
◦ between 0.3 and 1.5 THz
(< 0.1
◦ from 0.3 to 2.5 THz).
The above demonstrates the drawbacks in using WGPs in polarization-resolved
detection or in ellipsometric schemes to study anisotropic materials. While the above
WGP may not be competitive with some of the best commercial WGPs (which have
smaller periods), any WGP will introduce uncertainty into the identification of optical
properties. Comparatively, RP-THz-TDS does not suffer from this issue, and has an
accuracy and precision comparable to the best ellipsometric methods [6].
3.4 Experimental Implementation of RP-THz-TDS
To demonstrate how RP-THz-TDS can be utilised to investigate anisotropic media,
case studies were made of birefringence in two uniaxial crystals, ZnO and LaAlO 3 ,
and anisotropic absorption and birefringence in the biaxial crystal CuO. This section
will demonstrate how the RP-THz-TDS technique may be used to identify the orientations of polarisation eigenvectors and extract their complex optical properties,
and also how the orientations of anisotropic absorpion features may be investigated.
55
frequency-dependent elliptical pulse, which is approaching circular polarisation for
frequencies above 3.0 THz. This ellipticity occurs due to the finite transmission of the
component of the THz electric field parallel to the wires; this component is the first
time derivative of the incident field [19], and is therefore phase-shifted by a factor
of π/2 with respect to the orthogonal component perpendicular to the wires. The
orientation angle also deviates significantly from the intended angle of ψ = −35.0
◦ ,
by 2.0
◦ in the best case at low frequency, and with larger change as the transmitted
pulse becomes more elliptical.
As demonstrated in Fig. 3.6b, the WGP was then removed and the emitter rotated
to ψ in = −35.0
◦ , the same angle as that discovered to be perpendicular to the wires
of the WGP in the first section of this experiment. This simulates rotating the polarisation state by 45
◦ by rotating the emitter, allowing direct comparison of rotating
the polarisation state and projecting it to the same angle. The frequency dependence
of the ellipticity and orientation angle of THz pulses in this case is shown by the
red lines in Fig. 3.6c, d, respectively. The rotatable polarization method produces a
linear THz pulse over the whole experimental bandwidth when rotated to an arbitrary angle, with an ellipticity of ≤ 0.75
◦ and a statistical error (the precision) in the
ellipticity of < 0.05
◦ over the 0.3−1.5 THz range (< 0.1
◦ from 0.3 to 2.5 THz) after
20 repeated measurements. The accuracy of the orientation angle was defined as the
variation of ψ(ω) away from the mean [6], and was < 1.0
◦ between 0.3 and 2.5 THz.
The statistical error in the orientation angle was < 0.05
◦ between 0.3 and 1.5 THz
(< 0.1
◦ from 0.3 to 2.5 THz).
The above demonstrates the drawbacks in using WGPs in polarization-resolved
detection or in ellipsometric schemes to study anisotropic materials. While the above
WGP may not be competitive with some of the best commercial WGPs (which have
smaller periods), any WGP will introduce uncertainty into the identification of optical
properties. Comparatively, RP-THz-TDS does not suffer from this issue, and has an
accuracy and precision comparable to the best ellipsometric methods [6].
3.4 Experimental Implementation of RP-THz-TDS
To demonstrate how RP-THz-TDS can be utilised to investigate anisotropic media,
case studies were made of birefringence in two uniaxial crystals, ZnO and LaAlO 3 ,
and anisotropic absorption and birefringence in the biaxial crystal CuO. This section
will demonstrate how the RP-THz-TDS technique may be used to identify the orientations of polarisation eigenvectors and extract their complex optical properties,
and also how the orientations of anisotropic absorpion features may be investigated.