Speed of Response

The set-up used to measure the detector response to pulses from an excimer laser is shown schematically in Figure 3.10.

For the measurements carried out at 193 nm a Lambda Physik LPX210i excimer laser was used with an ArF lasing medium; the laser was filled with 170 mbar Ar, 80 mbar F] and 2750 mbar Ne. For the measurements carried out at 157 nm, reported in Chapter 7, the lasing medium was F2 gas in a modified Lambda Physik LPX210i excimer laser

The laser beam is guided using appropriate beam optics for the wavelength being used. Along the beam line were a shutter and a set of variable attenuator plates, allowing complete termination of the beam and the ability to reduce the power of the laser beam, respectively. A Uncoated glass p la te X ^ T V a r i a b l e / ^ i attenuator Beam ootics Shutte 'Detector bias

Device Digital storage

under test

'Vacuum photodiode Vacuum photodiode power supply

Pulsed excimer laser

Figure 3.10 Schematic diagram o f set-up used to measure the detector response to an excimer laser.

The laser beam has cross-sectional dimensions of about 15 x 25 mm. The device under test is held in the centre of this beam where the beam profile is reasonably flat. It is biased with a d.c. bias; in general a 9V PP3 battery was used to bias the device, but a power supply with variable d.c. voltage output could be used. The detector and bias were connected in series with the terminal of a digital storage oscilloscope; a termination of 50 Q was used on the oscilloscope. This results in a potential divider set­ up with the bias voltage being dropped across the detector and oscilloscope in proportion to their resistance values. The output is measured by the oscilloscope across its 50 Q termination. When not illuminated by the laser beam the diamond photoconductor has a resistance of >Gf], hence the entire bias voltage is dropped across the detector and the output at the oscilloscope is zero. If the detector is illuminated such that its resistance falls significantly a voltage output may be seen at the oscilloscope.

The oscilloscope used during these measurements was a Tektronix TDS3052 digital phosphor oscilloscope with a sampling rate of 5 giga-samples per second and a

Chapter 3 : Experimental Methods

frequency of 500 MHz. This oscilloscope can output the sampled data (either 500 or

1 0 , 0 0 0 data points) to a suitable file format on its internal floppy disk drive and the

screen image to either an image file or to a printer. Built-in functions enable automatic calculation of such pulse characteristics as peak height, full width at half maximum, and rise and fall times.

A ultra-violet vacuum photodiode (Instrument Technology Limited TF1850 M20UV K.3.3) was used to obtain a comparison for the temporal characteristics of the laser pulse. It was biased with 1.5 kV from a power supply. Due to the sensitivity of the vacuum photodiode it was not placed directly in the laser beam, but was illuminated with light reflected from an uncoated glass plate placed in the beam; approximately 4% of the pulse power is reflected in such a configuration [Rizvi, 1999].

Measurement of the laser fluence was performed by placing a calorimeter (Molectron Energy Max 400) in the beam line in front o f the device under test. This would measure the pulse energy so a mask with a 1 cm square aperture is placed in front of the

calorimeter, positioned so that the aperture would enable the device to be illuminated. The energy (in mJ) read on the calorimeter then corresponds to the fluence (in mJcm'^) of the laser pulse.

The lack of suitable attenuator plates for use at 157 nm, along with the strong attenuation of this wavelength in air, meant that for temporal measurements using the F2

excimer laser the experimental arrangement was slightly different; the beam was conveyed to the device under test (diamond photoconductor or vacuum photodiode) via a tube purged with nitrogen and terminated with a simple iris, and the beam power/fluence was controlled by varying the laser EHT.

The typical linewidth of an excimer laser is about 0.1 nm and pulse-to-pulse variation can be about ±5% for an ArF excimer and ±15% for an F2 excimer laser [Hecht, 1992].

Beams normally have flat or Gaussian spatial profiles; often they have a flat profile in one direction and a Gaussian profile in the orthogonal direction.

In an excimer laser the bulk of the mixture, 90 - 99%, is a buffer gas which is not part of the light emitting species but mediates energy transfer. Both helium and neon are used

rare-gas-halide laser, such as ArF, the rare gas is present in higher concentrations than the halogen. However, the gas mixture varies with use and with time, regardless of whether the laser is used, as the halogen concentration gradually decreases. Eventually the laser cavity must be pumped out and the laser gases replaced. This gas lifetime varies with type of laser gas, operating conditions and also with different laser models; Table 3.1 shows example gas lifetimes (number of shots) for one particular multigas excimer laser at different wavelength [Hecht, 1992]. High halogen concentrations extend gas lifetimes before replacement, but lower concentrations given more uniform beams. The kinetics of energy transfer in an excimer laser are complex and shall not be discussed here as this thesis is concerned with the detection of laser pulses, not the generation of them.

Laser W avelength (nm) Life (10^ shots)'

F2 157 0.05

ArF 193 0.4

KrF 248 1.0

XeCl 308 10.0

XeF 351 2.0

Table 3.1 Excimer gas lifetimes in a multigas laser without special cleaning equipm ent for gas replenishment. ‘Representative data; actual life depends on operating conditions.

Reproduced from [Hecht, 1992]

Air strongly absorbs light of wavelengths shorter than 200 nm (for this reason they are known as vacuum ultraviolet). The 193 nm light from an ArF excimer laser is attenuated about 10% per metre in air [Hecht, 1992], but 157 nm radiation from a F2

excimer laser is rapidly absorbed by air, hence the use of a nitrogen purged tube between the laser output and the device under test. Care must be taken when using such lasers because ultraviolet exposure can harm the skin and, even though wavelengths shorter than ~315 nm do not reach the retina, they are absorbed in the cornea where they can cause a painful but temporary sunbum-like effect, sometimes called "snow blindness" or "welder's flash". Absorption of VUV radiation in air can also create ozone which is harmful.

Other safety issues include the high voltages (EHT o f tens o f kilovolts) which are used to create the discharge within the laser, and the dangers of using fluorine. Rare gases, _ _

Chapter 3: Experimental Methods

such as Ar, are relatively safe, other than problems with high pressure cylinders and possible displacement of oxygen, but pure F2 gas is so corrosive that it presents fire and

explosion hazards. Safety standards limit fluorine content to 0.1 parts per million (ppm) in air [Hecht, 1992]. Hence, both the laser mixture and waste gas from the laser present hazards.