Time-Resolved Phosphorescence Detection (TRPD)
Time-resolved detection of O2(1Dg) is usually accomplished taking advantage of its weak phosphorescence at 1.27 mm. The intensity of this emission and therefore of the electrical signal produced by a detector is proportional to the O2(1Dg) concentration. For homogeneous systems:
The TRPD set-up is readily implemented in a laboratory already having pulsed lasers and consists of the following elements:
 | A pulsed laser with emission matching the absorption spectrum of the photosensitizer, with pulse duration in the nanosecond time-scale or lower, and ca. 1 mJ energy per pulse. |
| A suitable detector, sensitive in the near-IR region, with rise-time below ca. 1 ms. |
| A digital oscilloscope or transient recorder with digitizing frequency 10 MHz or higher |
| Optical elements such as neutral density filters or crossed-beam polarizers to attenuate the laser output, and cut-off and interference filters to filter out any sensitizer fluorescence and residual laser light in the detection port. |
The most common arrangement between the excitation and observation ports is the right-angle geometry. A standard fluorescence cell can be used for the sample. Placing the cell in a mirrored holder with just two holes drilled for entry and exit of the laser beam is advantageous in order to (1) increase the signals and (2) shield the detector from spurious light. Some authors prefer to use more sophisticated combinations of mirrors and lenses to collect as large a fraction as possible of the weak emission.11
Emission filters are used to prevent both scattered laser light and sensitizer fluorescence or phosphorescence from impinging onto the detector. Observation of large spikes in the early stages of the signal is one of the most common problems since they often produce saturation of the detector electronics and obscure the early stages of the signal.
Tip: Shielding the detector from stray light is an essential requirement for obtaining good kinetic traces. The simplest approach is to use a cut-off silicon filter, which blocks light below 1050 nm, in combination with an interference filter for selection of a narrow range around 1.27 mm. It is however known that silicon filters do emit light when illuminated with UV-Vis light.12 A simple trick is to use an array of long-pass filters with increasingly higher cut-off wavelengths.13 It is also essential to eliminate NIR background radiation, a problem often encountered when using a Nd:YAG laser. It is advisable to tape an IR-blocking filter (e.g. Schott KG5) at every port of the laser.The detector of choice for O2(1
Dg)-NIR emission is still the germanium photodiode. A number of manufactures supply the detector "ready to use", i.e., properly wired for time-resolved measurements, coupled to a preamplifier, and even encased in a cooling unit such as a Dewar flask for noise reduction.Alternative detectors replacing germanium photodiodes are InGaAs diodes.14 Hamamatsu has recently released a photomultiplier tube that will probably replace the photodiodes owing to its higher gain and faster response time.15 The availability of such tubes opens new possibilities to the field as they allow for time-correlated single-photon counting detection of O2(1
Dg) phosphorescence, an approach which has already been applied with custom-made tubes.16It s recommended to add a laser energy meter. This is especially important for determining
FD values and also to rule out undesired phenomena such as two-photon absorption or ground-state depletion.
Other spectroscopic techniques
In addition to monitoring the weak phosphorescence at 1270 nm, time-resolved detection of O2(1
Dg) can be accomplished also taking advantage of its emission of heat or of its absorption of light. | Time-resolved thermal lensing uses the local solvent refractive index changes caused by the release of heat accompanying O2(1Dg) formation and decay. This technique affords FD and tD values in a single experiment without the need for an external reference.17,18 However, it requires transparency of the solution. |
| Laser-induced optoacoustic spectroscopy is also a photothermal technique used for the determination of FD values.19,20 In this case, the pressure wave generated by the local solvent heating is monitored with a piezoelectric detector. |
| Time-resolved O2(1Dg) absorption (in the IR) is the latest addition to this field.21 |
Sample preparation and handling
Solvents should be purified following standard procedures.22 The value of
tD is especially sensitive to impurities containing OH groups.23 Hence, every effort must be done to eliminate residual water from the solvent of choice. Solutions should always be prepared immediately before use and handled under dim light. Storing solutions over long periods of time is discouraged.It is often desirable to adjust the O2 content of the solutions by bubbling with mixtures containing an inert gas such as nitrogen, helium, or argon. A convenient way of achieving this is by the use of flow meters to adjust the proportions of the two gases. The gases should always be flown through at least two beakers containing the desired solvent for minimizing solvent evaporation in the sample.
Recording O2(1Dg) kinetic traces
The diode signal should be coupled to the digitizer through a 50
W entrance resistor for better time resolution, always paying attention to achieving good impedance matching. Impedance mismatch often results in signals "going below zero". The time-base in the digitizer should be chosen such that 4-5 lifetimes are recorded. The signal should be authenticated to rule out artifacts or spurious emissions: it should disappear when oxygen is excluded from the system (e.g., upon inert gas saturation) and it should be possible to reduce its lifetime (not its amplitude) adding well-established O2(1Dg) quenchers such as azide, 1,4-diazabicyclo[2.2.2]octane (DABCO), or b-carotene.24