3.2 Rotatable-Polarisation Terahertz Time-Domain Spectrometer
49
and the x-axis of the lab coordinate system, and thus the orientation which defines
ψ = 0, is defined by the orientation of the [211]-axis of the ZnTe detection crystal.
Both the QWP and HWP were mounted in high-precision motorised rotation stages
to facilitate repeatable and remotely-controllable rotation of these optics.
The initial step in the alignment procedure is to define δ = 0; since the output of
the laser and therefore the gate beam is horizontally polarised, this is achieved by
rotating the HWP to find the orientation θ
V
HWP which minimises the intensity of the
vertically-polarised arm of the gate beam after the WP, without the QWP present
in the gate beam path. The next step was to place the QWP and ZnTe into the gate
beam path, and rotate the QWP to find the angle θ
V
QWP at which the output of the
photodiode circuit is balanced. The rotatable emitter was then set at ψ em = 0, and the
ZnTe crystal was rotated about its surface normal to find the maximum electro-optic
signal, which defines θ = 0. By referring to Eq. 2.6 and Fig. 2.3, we see that when
both δ = 0 and θ = 0 the detection system is only sensitive to the E 011 component of
the THz electric field, corresponding to vertically polarised THz radiation parallel to
the lab y-axis, E y . The x component of the THz radiation, E x , was then obtained by
setting δ = 22.5
◦ by precisely rotating the HWP to θ
H
HWP , and the photodiode output
was balanced again by rotating the QWP to θ
H
QWP .
This alignment procedure therefore allows the detection sensitivity to be easily
switched between two orthogonal components of the THz pulse; by rotating the HWP
and QWP to known values of θ
V
HWP and θ
V
QWP to detect the vertical component, and
to θ
H
HWP and θ
H
QWP to detect the horizontal component.
3.2.3 Calibration of the Terahertz Emission Strength
As defined in Sect. 3.1.3, an ideal polarisation rotation system should have a minimal
variation in the amplitude of the electric field |E| =
E 2
x + E 2
y . Therefore in the RPTHz-TDS system we want to achieve a minimal variation in |E| with ψ em , in order to
achieve the most consistent operation of the spectrometer possible for all orientation
angles. On initially rotating the emitter it was apparent that its center was close to,
but not exactly on, the axis of the rotation stage. Thus under rotation the emitter was
no longer centered on the THz generation beam, altering the radiated THz power. To
overcome this problem the centre of the emitter was found at a range of different ψ em
from ψ em = 0 to ψ em = 180
◦ , which allowed us to re-centre and photoexcite the same
area on the emitter at each angle. The rotation stage holding the emitter was mounted
on a motorized x-y translation stage to facilitate precicely controllable adjustment
of the emitter position in the x y-plane. The device was then raster scanned, under
illumination by the photoexcitation beam, and the device resistance was measured
at each ψ em in steps of 10
◦ . When the photoexcitation beam is incident on the active
area of the device, the generated photocurrent causes the resistance of the device to
decrease; in comparison when the beam is incident on an area of the substrate outside
of the active area, the resistance will remain at a higher value. Example maps at a few
49
and the x-axis of the lab coordinate system, and thus the orientation which defines
ψ = 0, is defined by the orientation of the [211]-axis of the ZnTe detection crystal.
Both the QWP and HWP were mounted in high-precision motorised rotation stages
to facilitate repeatable and remotely-controllable rotation of these optics.
The initial step in the alignment procedure is to define δ = 0; since the output of
the laser and therefore the gate beam is horizontally polarised, this is achieved by
rotating the HWP to find the orientation θ
V
HWP which minimises the intensity of the
vertically-polarised arm of the gate beam after the WP, without the QWP present
in the gate beam path. The next step was to place the QWP and ZnTe into the gate
beam path, and rotate the QWP to find the angle θ
V
QWP at which the output of the
photodiode circuit is balanced. The rotatable emitter was then set at ψ em = 0, and the
ZnTe crystal was rotated about its surface normal to find the maximum electro-optic
signal, which defines θ = 0. By referring to Eq. 2.6 and Fig. 2.3, we see that when
both δ = 0 and θ = 0 the detection system is only sensitive to the E 011 component of
the THz electric field, corresponding to vertically polarised THz radiation parallel to
the lab y-axis, E y . The x component of the THz radiation, E x , was then obtained by
setting δ = 22.5
◦ by precisely rotating the HWP to θ
H
HWP , and the photodiode output
was balanced again by rotating the QWP to θ
H
QWP .
This alignment procedure therefore allows the detection sensitivity to be easily
switched between two orthogonal components of the THz pulse; by rotating the HWP
and QWP to known values of θ
V
HWP and θ
V
QWP to detect the vertical component, and
to θ
H
HWP and θ
H
QWP to detect the horizontal component.
3.2.3 Calibration of the Terahertz Emission Strength
As defined in Sect. 3.1.3, an ideal polarisation rotation system should have a minimal
variation in the amplitude of the electric field |E| =
E 2
x + E 2
y . Therefore in the RPTHz-TDS system we want to achieve a minimal variation in |E| with ψ em , in order to
achieve the most consistent operation of the spectrometer possible for all orientation
angles. On initially rotating the emitter it was apparent that its center was close to,
but not exactly on, the axis of the rotation stage. Thus under rotation the emitter was
no longer centered on the THz generation beam, altering the radiated THz power. To
overcome this problem the centre of the emitter was found at a range of different ψ em
from ψ em = 0 to ψ em = 180
◦ , which allowed us to re-centre and photoexcite the same
area on the emitter at each angle. The rotation stage holding the emitter was mounted
on a motorized x-y translation stage to facilitate precicely controllable adjustment
of the emitter position in the x y-plane. The device was then raster scanned, under
illumination by the photoexcitation beam, and the device resistance was measured
at each ψ em in steps of 10
◦ . When the photoexcitation beam is incident on the active
area of the device, the generated photocurrent causes the resistance of the device to
decrease; in comparison when the beam is incident on an area of the substrate outside
of the active area, the resistance will remain at a higher value. Example maps at a few