LUMINESCENCE SPECTROSCOPY APPARATUS

Abstract

A luminescence spectroscopy apparatus for time-resolved characterization of a sample emitting circularly polarized light is described, comprising a pulsed laser excitation source configured to generate a laser pulse for exciting the sample, an achromatic quarter-wave plate arranged to receive therethrough light emitted by the sample, a polarization beam splitter arranged downstream to the quarter-wave plate, an optical spectrometer arranged downstream to the polarization beam splitter, a time-gated intensified charge-coupled device arranged to receive light from the optical spectrometer and comprising an image intensifier, and a controller configured to control a pulse generator to apply a gate pulse to the image intensifier for selectively activating the image intensifier, wherein the controller is connected to the pulsed laser excitation source such that the gate pulse is triggerable by the laser pulse of the pulsed laser excitation source.

Claims

1. A luminescence spectroscopy apparatus for time-resolved characterization of a sample emitting circularly polarized light, comprising a pulsed laser excitation source configured to generate a laser pulse for exciting the sample, an achromatic quarter-wave plate arranged to receive therethrough light emitted by the sample, a polarization beam splitter arranged downstream to the quarter-wave plate, an optical spectrometer arranged downstream to the polarization beam splitter, a time-gated intensified charge-coupled device arranged to receive light from the optical spectrometer and comprising an image intensifier, and a controller configured to control a pulse generator to apply a gate pulse to the image intensifier for selectively activating the image intensifier, wherein the controller is connected to the pulsed laser excitation source such that the gate pulse is triggerable by the laser pulse of the pulsed laser excitation source.

2. The luminescence spectroscopy apparatus according to claim 1, wherein the polarization beam splitter is a birefringent polarization beam splitter.

3. The luminescence spectroscopy apparatus according to claim 2, wherein the birefringent polarization beam splitter is a Wollaston polarizer or a Rochon polarizer.

4. The luminescence spectroscopy apparatus according to claim 1, comprising an achromatic half-wave arranged in sequence with the quarter-wave plate.

5. The luminescence spectroscopy apparatus according to claim 4, wherein the quarter-wave plate is arranged downstream to the half-wave plate.

6. The luminescence spectroscopy apparatus according to claim 1, wherein the charge-coupled device comprises an image area with a first region of interest configured to receive light with a first polarization component and a second region of interest configured to receive light with a second polarization component.

7. The luminescence spectroscopy apparatus according to claim 1, comprising a first motor controller, wherein the quarter-wave plate is rotatably mounted in a motorized rotary mount, wherein the first motor controller is configured to control rotation of the quarter-wave plate.

8. The luminescence spectroscopy apparatus according to claim 1, comprising a second motor controller, wherein the half-wave plate is rotatably mounted in a motorized rotary mount, wherein the second motor controller is configured to control the rotation of the half-wave plate.

9. The luminescence spectroscopy apparatus according to claim 1, wherein the quarter-wave plate and the polarization beam splitter define a detection path downstream of the sample, wherein the pulsed laser excitation source is configured to define an excitation path which is perpendicular to the detection path.

10. The luminescence spectroscopy apparatus according to claim 9, wherein the pulsed laser excitation source is configured to generate a laser pulse having a horizontal polarization.

11. The luminescence spectroscopy apparatus according to claim wherein the quarter-wave plate and the polarization beam splitter define a detection path downstream of the sample, wherein the pulsed laser excitation source is configured to define an excitation path which is parallel to the detection path.

12. A method of time-resolved characterization of a sample emitting circularly polarized light, comprising the steps of: a) providing a luminescence spectroscopy apparatus according to claim 1; b) generating by the pulsed laser excitation source a laser pulse for exciting a sample; c) receiving by the controller a trigger signal from the pulsed laser excitation source; d) triggering the pulse generator by the controller to generate a gate pulse; e) activating the image intensifier by applying the gate pulse by the pulse generator to the image intensifier; f) recording simultaneously a first polarization component of light emitted by the sample on a first region of interest of an image area of the charge-coupled device and a second component of light emitted by the sample on a second region of interest of the image area of the charge-coupled device.

13. The method according to claim 12, comprising the steps of: executing the steps of b)-f) at a first orientation of the quarter-wave plate with its fast axis being at a first quarter-wave plate angle; rotating the quarter-wave plate by a first motor controller to a second orientation with the fast axis being at a second quarter-wave plate angle, wherein the first quarter-wave plate angle and the second quarter-wave plate angle differ by 90; repeating the steps of b)-e), and recording the first polarization component of light emitted by the sample on the second region of interest of the image area of the charge-coupled device and the second component of light emitted by the sample on the first region of interest of the image area of the charge-coupled device.

14. The method of claim 13, wherein after rotating the quarter-wave plate by the first motor controller to the second orientation, beam steering is executed on the pulsed laser excitation source until the light from the pulsed laser excitation source is collected by the same pixels of the charge-coupled device as before rotation of the quarter-wave plate.

15. The method according to claim 13, comprising the steps of: rotating the quarter-wave plate by the first motor controller to a third orientation with the fast axis being at a third quarter-wave plate angle, wherein the second quarter-wave plate angle and the third quarter-wave plate angle differ by 90; repeating the steps of b)-f).

16. The method according to claim 15, comprising the steps of: rotating the quarter-wave plate by the first motor controller to a fourth orientation with the fast axis being at a fourth quarter-wave plate angle, wherein the third quarter-wave plate angle and the fourth quarter-wave plate angle differ by 90; repeating the steps of b)-e); recording the first polarization component of light emitted by the sample on the second region of interest of the image area of the charge-coupled device and the second component of light emitted by the sample on the first region of interest of the image area of the charge-coupled device.

17. The method according to claim 12, comprising the steps of: providing in step a) an achromatic half-wave plate arranged in sequence with the quarter-wave plate; executing the steps of b)-f) at a first orientation of the half-wave plate with its fast axis being at a first half-wave plate angle; rotating the half-wave plate by a second motor controller to a second orientation with the fast axis being at a second half-wave plate angle, wherein the first half-wave plate angle and the second half-wave plate second angle differ by 45; repeating the steps of b)-e), and recording the first polarization component of light emitted by the sample on the second region of interest of the image area of the charge-coupled device and the second component of light emitted by the sample on the first region of interest of the image area of the charge-coupled device.

18. The method of claim 17, wherein after rotating the half-wave plate by the second motor controller to the second orientation, beam steering is executed on the pulsed laser excitation source until the light from the pulsed laser excitation source is collected by the same pixels of the charge-coupled device as before rotation of the half-wave plate.

19. The method according to claim 12, wherein vertical pixel binning is executed in step f) such that only a first track of the first polarization component and a second track of the second polarization component is output from the charge-coupled device.

20. The method according to claim 12, wherein the steps b)-f) are repeated by incrementally changing a position of a grating of the optical spectrometer after each step f).

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0060] The present invention will be explained in more detail, by way of exemplary embodiments, with reference to the schematic drawings, in which:

[0061] FIG. 1 shows a perspective view of an embodiment of a luminescence spectroscopy apparatus;

[0062] FIG. 2 shows a block diagram of an embodiment of a luminescence spectroscopy apparatus;

[0063] FIG. 3 shows an illustration of the gate pulses for time-gating;

[0064] FIG. 4 shows a simulated example illustrating combining successive measurements by waveplate rotation and simultaneous measurements by two-channel detection;

[0065] FIG. 5 shows calculated intensity differences between measurement tracks as a function of both QWP and HWP angles for the pure Stokes basis polarizations;

[0066] FIG. 6 shows steady-state and microsecond time-resolved CPL spectroscopy of the chiral standard Eu[(+)-facam].sub.3;

[0067] FIG. 7 shows measurements of nanosecond time-resolved non-polarized, circular and photoselection-induced linear anisotropy of Eu[(+)-facam].sub.3;

[0068] FIG. 8 shows enantiomer CPL comparison and time-resolved CPL for the R-enantiomer of a chiral TADF-active compound;

[0069] FIG. 9 shows broadband steady-state and time-resolved full Stokes-vector spectroscopy of a standard achiral dye in water-based solutions with low and high viscosity.

DETAILED DESCRIPTION OF EXEMPLARY EMBODIMENTS

[0070] FIG. 1 shows a perspective view of an embodiment of a luminescence spectroscopy apparatus 10 comprising a pulsed laser excitation source 1. In FIG. 1, the pulsed laser excitation source 1 is shown twice in order to illustrate two configurations of the excitation path. In the first configuration realizing an orthogonal geometry, the pulsed laser excitation source 1 is arranged such that the excitation path along which the laser pulse P1 propagates is perpendicular to the detection path along which the light L emitted by the sample S propagates. In the second configuration realizing a parallel geometry, the pulsed laser excitation source 1 is arranged such that the excitation path along which the laser pulse P2 propagates is parallel to the detection path along which the light L emitted by the sample S propagates.

[0071] The laser pulse P1 or P2 excites the sample S which emits light L that propagates through a collimating lens 1.1 and reaches an achromatic half-wave plate 2 by means of which the linearly polarized components of the light L can be quantified. An achromatic quarter-wave plate 3 is arranged downstream to the half-wave plate 2 and receives the light that has passed therethrough. The half-wave plate 2 and the quarter-wave plate 3 are each rotatably arranged on a motorized rotary mount and can be rotated into appropriate orientations to transform polarization components into orthogonal linear components, which are separated by a Wollaston prism 4 arranged downstream to the quarter-wave plate 3. After passing a focusing lens 1.2 and a filter 1.3, the separated components pass through a grating spectrometer 5. The orthogonally polarized spectra are then simultaneously recorded as two tracks v and h on a single CCD array 62 of a CCD detector 6. The intensified CCD detector further comprises an intensifier 61 which is electronically gated to provide time resolution. A pulse generator 71 is configured to apply a gate pulse having a certain delay and width to the intensifier 61 upon being triggered by a controller 72. The controller 72 receives a trigger signal from the pulsed laser excitation source 1 upon generation of a laser pulse P1 or P2 for exciting the sample S.

[0072] FIG. 2 shows a block diagram of an embodiment of a luminescence spectroscopy apparatus 10. The pulsed laser excitation source 1 comprises a pulsed laser 11 and excitation optics 12 with optical components such as mirrors, lenses, filters, polarizers etc. to define the excitation beam. In particular, the excitation optics 12 comprises polarizers 13 to define the desired polarization of the excitation laser pulse. Using the excitation optics 12, the excitation path P1 or P2 can be defined. Furthermore, the excitation optics 12 comprises a steering mirror 14 by which beam steering can be executed in order to correct for beam drift effects.

[0073] The light L emitted by the source S passes through a superachromatic half-wave plate 2, a superachromatic quarter-wave plate 3 and a Wollaston prism 4 which separates the orthogonal polarization components, which then propagate to the spectrograph 5. The two polarization tracks are collected by the iCCD detector 6. The controller 72 controls the iCCD detector 6, the spectrograph and the first motor controller 31 and the second motor controller 21. The first motor controller 31 and the second motor controller 21 are used to rotate the quarter-wave plate 3 and the half-wave plate 2, respectively, which are each mounted on a motorized rotary mount. The intensifier of the iCCD detector 6 is triggered by a trigger signal from the pulsed laser 11. The controller 72 is used for simultaneous control of the components of the luminescence spectroscopy apparatus 10 such that an automated TRCPL measurement and full Stokes measurement can be provided.

Example Luminescence Spectroscopy Setup

[0074] In the following, an example luminescence spectroscopy setup is described.

[0075] The samples are optically excited using the output of a Light Conversion PHAROS laser (Yb:KGW lasing medium, 1030 nm, pulse energy 400 J, pulse width duration 200 fs, repetition rate of 50 kHz). The pump beam is generated from the seed in a harmonic generation (Light unit Conversion HIRO) via nonlinear crystals (beta-barium borate, lithium triborate) with residual fundamental removed by dichroic mirrors within the unit. Second and third harmonics can be generated, giving pump wavelengths of 515 nm or 343 nm respectively. Pump pulse energy at the sample is 10-70 nJ, with the pump focused down to a beam diameter of approximately 1 mm. The laser repetition rate is controllable by a pulse picker, and repetition rates in the range 500 Hz-50 kHz are used.

[0076] The CCD sensor has 2048512 pixels, enabling simultaneous recording of multiple tracks. Horizontal and vertical polarization components, which are spatially separated by a Wollaston prism (shown on the sensor in Figure S2), can therefore be simultaneously recorded. Vertical pixel binning is used to produce two effective vertical pixels for each wavelength pixel, giving I.sub.h() and I.sub.v() for the horizontal and vertical channels respectively.

[0077] For time-resolved measurements, the quantities recorded during a single acquisition are I.sub.h(, t) and I.sub.v(, t) where t is defined by the gate pulse applied, as illustrated in FIG. 3. The sample is excited by a laser pulse arriving at t=0, with the laser sync signal triggering a gate pulse on the detector with a user-defined delay and width. The detector only collects light when the gate pulse is on, resulting in collection of time-resolved spectra. Time series are built up by repeating the measurement with modified gate delays/widths.

[0078] Polarization tracks are separated from a single beam by a Wollaston prism which remains fixed. Therefore, one track will correspond to a vertical polarization and the other to a horizontal polarization. Waveplates are used to make the polarization components of interest (0/90, +45/45, LCP/RCP) fall completely to the horizontal or vertical tracks, and then another waveplate orientation is used to swap the components on the tracks.

[0079] This swapping of tracks, combined with measuring both tracks at two such waveplate orientations, allows for cancelling out the bulk of errors caused by any transmission inequalities between the two beam paths. Further, the simultaneous recording of both tracks cancels out the bulk of time-instability errors, arising from e.g. drift in excitation intensity, which is illustrated in FIG. 4. FIG. 4 shows how combining successive measurements by waveplate rotation and simultaneous measurements by two-channel detection can accurately recover all Stokes components with even substantial spatial and temporal error sources. Here, track transmission mismatch has been set to 20% and a random time drift set to 5% of the incident light intensity.

[0080] After the Wollaston prism, the channels are horizontally and vertically polarized (0/90 respectively), corresponding to the S.sub.1 polarization axes. To investigate S.sub.2 (linear)+45 and S.sub.3 (LCP/RCP) polarization components, waveplates are used to transform these polarization components to 0/90 linear polarizations. For S.sub.2, suitable polarization transformation is achieved by a half-wave plate (HWP) at 22.5+45, and for S.sub.3 with a quarter-wave plate (QWP) at 45+90, where the waveplate orientations are defined about the fast axis relative to the table plane. Any two adjacent orientations will swap the polarization component imaged on a given track; for example, if the RCP component is measured at the horizontal track for a 45 QWP orientation, at a 135 QWP orientation the RCP component is measured at the vertical track.

[0081] For a fast, automatable full-Stokes measurement, it is advantageous to have both waveplates in place for all measurements. An illustration of how the various Stokes components produce intensity differences between the tracks in the ideal case of perfect waveplates with both a HWP and QWP present is presented in FIG. 5. For CPL measurements, the HWP orientation in principle does not matter, but in all orientations the HWP will act to transform LCP to RCP and vice versa. The waveplate angles for measuring a given Sn component are such that the other two components are split evenly across the two channels and therefore are not measured as a false polarization signal.

[0082] As described above, two tracks are simultaneously recorded at a given orientation of the HWP and QWP. The recorded intensities are denoted as I.sub.v,QwP,HWP and I.sub.h,QwP,HWP.

[0083] An S.sub.3 measurement would then seek to find I.sub.LCP and I.sub.RCP, recorded as

[00001]$I_{\mathrm{LCP}}=I_{h,\mathrm{QWP}45,\mathrm{HWP}0}+I_{v,\mathrm{QWP}135,\mathrm{HWP}0}$$I_{\mathrm{RCP}}=I_{v,\mathrm{QWP}45,\mathrm{HWP}0}+I_{h,\mathrm{QWP}135,\mathrm{HWP}0}$

[0084] From which the following quantities are calculated

[00002]$I=I_{\mathrm{LCP}}-I_{\mathrm{RCP}}$$I_{\mathrm{total}}=I_{\mathrm{LCP}}+I_{\mathrm{RCP}}$

[0085] Whereas for an S.sub.1 measurement one has

[00003]$I_{0^0}=I_{h,\mathrm{QWP}0,\mathrm{HWP}0}+I_{v,\mathrm{QWP}0,\mathrm{HWP}45}$$I_{{90}^0}=I_{v,\mathrm{QWP}0,\mathrm{HWP}0}+I_{h,\mathrm{QWP}0,\mathrm{HWP}45}$$I=I_{0^0}-I_{{90}^0}$

[0086] And for an S.sub.2 measurement

[00004]$I_{+{45}^0}=I_{h,\mathrm{QWP}0,\mathrm{HWP}22.5}+I_{v,\mathrm{QWP}0,\mathrm{HWP}67.5}$$I_{-{45}^0}=I_{v,\mathrm{QWP}0,\mathrm{HWP}22.5}+I_{h,\mathrm{QWP}0,\mathrm{HWP}67.5}$$I=I_{+{45}^0}-I_{-{45}^0}$

[0087] The Stokes components are defined as

[00005]$S_0=I_0+I_{90}=I_{+45}+I_{-45}=I_{\mathrm{LCP}}+I_{\mathrm{RCP}}$$S_1=I_0-I_{90}$$S_2=I_{+45}-I_{-45}$$S_3=I_{\mathrm{LCP}}-I_{\mathrm{RCP}}$

[0088] Since the absolute count numbers are generally not of interest, the data can be scaled for convenience such that S.sub.0=1.

[0089] The linear polarization anisotropy (r) and circular polarization dissymmetry of the luminescence (g.sub.lum) are defined as

[00006]$r=\frac{I_{.Math.}-I_{}}{I_{.Math.}+2I_{}}$

[0090] where I.sub. and I.sub. are parallel and perpendicular linearly polarized intensities, and

[00007]$g_{\mathrm{lum}}=\frac{I_{\mathrm{LCP}}-I_{\mathrm{RCP}}}{I_{\mathrm{avg}}}=\frac{I_{\mathrm{LCP}}-I_{\mathrm{RCP}}}{\frac{1}{2}\left(I_{\mathrm{LCP}}+I_{\mathrm{RCP}}\right)}$

[0091] Electronic gating in the example setup has a minimum gate width of approximately 2 ns, which is also the overall limit of time resolution, as the gate delay accuracy and laser pulse width are far smaller.

[0092] Maximum time range is practically limited by the laser repetition rate, as very long gate pulse values are achievable. In the example setup, the laser operates at 50 kHz and can be pulse-picked for a lower frequency operation at the same per-pulse (and lower time-averaged power).

[0093] Lower repetition rates usually lead to slower data acquisition. Presently, the practical limit to perform polarization measurements is at approximately 500 Hz, corresponding to a time range of 2 ms; extending this could be done by increasing per-pulse power at the sample. Overall, accessible time bins range from approximately 2 ns-2 ms.

[0094] For accurate quantification of g.sub.lum, a large number of photons at each wavelength should be counted. As a rough estimate, reaching a noise level of 10.sup.4 requires 10.sup.8 photons in the ideal case where the only noise source is shot noise inherent to photon emission following Poisson statistics and signal-to-noise ratio scales with {square root over (N)} for N photons counted. In practice, this will be an underestimate for the example setup, as the intensified CCD detector has additional noise contributions compared to a photon-counting system (readout noise, shot noise in the dark signal, shot noise in the signal). It is found that measuring such a large number of photons tends to be relatively time-consuming, owing to limits on detector readout rate, laser repetition rate and detector saturation effects.

[0095] For faster measurements, the number of charges collected and read out per second is desired to be maximized. This is affected by excitation power (per pulse). Increasing the excitation laser (and thus gate pulse) repetition rate will increase signal strength, but normally it is desired to avoid wrap-around emission (i.e. emission that leaks into the photon collection after the next pulse), if possible. Beyond this, appropriate choice of time gates, exposure time and accumulations may be important. Where appropriate, pixel binning is a possible approach. For the present purposes, vertical pixels are binned such that only two tracks are recorded and read out, which greatly increases detector readout rate compared to a full-sensor readout.

[0096] Another consequence of using a multichannel detector such as a CCD is that there may be fixed pattern noise, that is, differences in count values read out by individual pixels when an identical amount of light is incident. This will be partially due to per-pixel variation of the dark signal, which will be temperature-dependent but otherwise fixed, and partially due to per-pixel variation in sensitivity due to manufacturing irregularities. The dark signal variation can be removed by background subtraction, and the per-pixel sensitivity variation, like other channel-dependent sensitivity variations, may be corrected for in the difference spectra and dissymmetry factor by the two-step measurement procedure where a QWP rotation flips the horizontal/vertical channels.

[0097] Rotation of the QWP may introduce a slight deflection of the beam. As such, the pixels which collect light in the second measurement are not exactly the same pixels as in the first measurement. Due to this, the track-swapping pixel sensitivity correction is incomplete, which may result in unexpectedly high static noise (i.e. consistent between measurements, assuming optics are not moved) approximately at the 10.sup.2 level. This can be corrected for by steering the excitation beam vertically after QWP rotation such that the same pixels are used to collect signal in both measurements. This can be facilitated by the detector having a live imaging readout which can be used for alignment. Beam steering or resteering, respectively, can be achieved by a steering mirror placed immediately before the sample. For the same measurement parameters, the apparent noise level is massively improved by manually resteering the beam after waveplate rotation. This is because the noise without beam resteering s not true statistical noise, but rather static imperfections in the error cancellation caused by changing the pixels over which data is collected.

[0098] Horizontal beam deflection (orthogonal to the spectrograph entrance slit) is largely mitigated by using a 90 excitation-collection geometry a with slow excitation focusing lens, such that the excitation beam forms a line through the sample. This has the added advantage of making alignment generally easier.

[0099] Beam drift introduced by waveplate rotation can be mitigated by performing measurements with more than the minimum of two waveplate orientations. For CPL measurements, a total of four different orientations can be used, and for linear polarization measurements eight orientations are possible. Using multiple orientations helps due to the beam drift being at least partially reversed upon further rotation of the waveplate (as rotation by 360 should return the beam to its initial position).

[0100] A further automatable approach to mitigate pixel sensitivity is the scanning multichannel approach. Here, instead of doing a single series of accumulations at a given grating position, a smaller number of accumulations is performed at multiple slightly offset grating positions. As an added benefit, this will similarly smooth out any larger-scale sensitivity variations of the intensifier/detector, which normally have slightly lower sensitivity near the edge regions, without requiring a manual calibration file.

[0101] Slit width introduces s a tradeoff between intensity and resolution. For a narrow-line emission like that of Eu.sup.3+, narrow (10-50 m) slit widths can be used to avoid smearing out features. For a pixel array detector, the pixel size usually introduces the lower limit to which slit size can improve resolution. Finer gratings allow for greater resolution at the cost of narrower bandpass (and, for holographic gratings, greater polarization sensitivity). For broader spectral shapes, such as the organic molecules investigated herein, a wider (100-200 m) slit allows for more light incoupling and is therefore advantageous.

[0102] A flow cell may be appropriate for sensitive applications, in particular, if sample degradation is an issue.

[0103] To obtain the static stokes vector, six measurements are to be executed in total. First, the measurement of S.sub.1 is carried out with the QWP fixed at 0, while the HWP is set to 0 and 45. Second, the measurement of S.sub.2 is performed with the QWP at 0, and the HWP rotated to 22.5 and 67.5. Third, the S.sub.3 (=g.sub.lum*S.sub.0) is performed with the HWP set at an arbitrary angle, along with the QWP orientations at 45 and 45. Executing these measurements effectively switches the polarization tracks and minimizes the contributions from other polarizations, achieving the desired results. To extend the steady-state Stokes vector with time resolution, additionally, each of the six measurements are performed at different time steps. These time steps are defined in terms of gate width and delay.

Example Measurements

[0104] In the following, example measurements using an embodiment of a luminescence spectroscopy apparatus are described.

Chiral Lanthanide Complex

Steady State and Microsecond Gating of Eu[(+)-Facam].SUB.3

[0105] Organometallic lanthanide complexes with chiral ligands are standard materials for strong CPL activity. In particular, Eu[(+)-facam].sub.3 ((+)-facam=3-(trifluoromethylhydroxymethylene)-(+)-camphorate) possesses strong emission dissymmetry (up to g.sub.lum=0.78 in DMSO at =595 nm). Thus, Eu[(+)-facam].sub.3 is a common choice for CPL setup validation. Accordingly, first steady-state and long timescale time-resolved (50 s gate width) CPL spectra of Eu[(+)-facam].sub.3 in dry DMSO are presented, as seen in FIG. 6. The features seen arise from f-f transitions on the Eu.sup.3+ center, which generally are of the form .sup.5D.sub.J1.fwdarw..sup.7F.sub.J2, more specifically here the features are assigned to .sup.5D.sub.0.fwdarw..sup.7F.sub.J2 transitions. Of these, the most striking CPL features are at 595 nm and 613 nm, respectively corresponding to an .sup.5D.sub.0.fwdarw..sup.7F.sub.1 magnetic dipole transition and .sup.5D.sub.0.fwdarw..sup.7F.sub.2 induced electric dipole transition, with literature g.sub.lum=0.78 and +0.072, respectively (albeit with some variance and environmental sensitivity, notably to water). In the steady state, our results match literature well. Time-resolved spectra (FIG. 6D-F) show little spectral shape evolution over time, but the measured g.sub.lum varies with time across the decay (FIG. 6G). The reduction in g.sub.lum over time is similar to what was observed before, where a decrease from a peak value of approximately 0.8 with a time constant of approximately 1 ms is reported and assigned to sample heterogeneity. Intriguingly, the data appears to show an increase in g.sub.lum over time for approximately the first 150 s; while the earliest data point in FIG. 6G may contain an element of instrument response, the rise continues beyond this. Therefore, both in the steady state and long timescale time-resolved measurements, the setup gives results consistent with previous literature for Eu[(+)-facam].sub.3. FIG. 6A shows total steady-state luminescence intensity (405 nm CW excitation), summed across all measurements. FIG. 6B shows steady-state luminescence intensity difference, representing excess counts of a given handedness, normalized by the maximum total luminescence intensity. FIG. 6C shows the steady-state dissymmetry factor g.sub.lum. FIG. 6D shows Total time-resolved luminescence spectra (343 nm excitation, 200 fs pulses at 1 kHz), normalized to the earliest time delay, 50 s bins. FIG. 6E shows time-resolved difference counts, normalized to the earliest bin. FIG. 6F shows the time-resolved dissymmetry factor. FIG. 6G shows intensity and dissymmetry factors as a function of time in a 0.3 nm range around the dissymmetry minimum.

Nanosecond Gating of Eu[(+)-Facam].SUB.3

[0106] Having shown measurement for the steady-state and long timescale TRCPL of Eu[(+)-facam].sub.3, short timescale TRCPL with 2-5 ns gating is shown in FIG. 7, being the first such results collected. FIG. 7A shows non-polarization sensitive luminescence spectra (excitation 343 nm, 200 fs, 500 Hz) with varying time bins spanning different luminescence regimes of Eu(facam).sub.3, normalized to highlight spectral changes. Key transitions around 400-500 nm (.sup.1LC), 550 nm (.sup.5D.sub.1.fwdarw..sup.7F.sub.2) and 595 nm (.sup.5D.sub.0.fwdarw..sup.7F.sub.1) are labelled. FIG. 7B shows non-polarization sensitive PL decay kinetic, integrated over the emission time-resolved CPL measurement spectrum. FIGS. 7C-E show (excitation 343 nm, 200 fs, 50 kHz, horizontal polarization) with 5 ns bins, using horizontal excitation polarization to avoid photoselection. FIGS. 7F-H show time-resolved horizontal/vertical linear polarization measurement (excitation 343 nm, 200 fs, 50 kHz, vertical polarization) with 2 ns bins and 0.5 ns steps, using vertical excitation polarization to intentionally induce photoselection.

[0107] Overall time-resolved spectra at various timescales are plotted in FIG. 7A), with a PL kinetic shown in FIG. 7B. Compared to steady state and long timescale spectra, a number of features besides the .sup.5D.sub.0.fwdarw..sup.7F.sub.J2 transitions are visible at early times, most obviously a broad, structureless peak around 435 nm and a series of narrower peaks in the 520-570 nm range.

[0108] For interpreting these additional features in the early-time Eu[(+)-facam].sub.3 spectra, it is useful to refer to the spectra of Eu.sup.3+ more generally. Eu.sup.3+ has a multitude of well-studied spectral lines arising from .sup.5D.sub.J1.fwdarw..sup.7F.sub.J2 transitions. As these are weakly absorbing (being weak magnetic dipole transitions or Laporte forbidden electric dipole transitions), PL intensity can be greatly enhanced by exciting symmetry-allowed electric dipole transitions involving a ligand. The ligands effectively act as antennae, transferring energy to the Eu.sup.3+ center from which subsequent emission occurs. The initially excited transition is the ligand singlet .sup.1LC, and energy transfer to Eu.sup.3+ occurs from the ligand spin-triplet .sup.3LC state. Therefore, intersystem crossing (ISC) must occur, with incomplete ISC or energy transfer resulting in ligand-centered emission. Energy transfer occurs preferentially via the .sup.5D.sub.1 level, per the selection rules for Dexter transfer. Emission from .sup.5D.sub.1, .sup.5D.sub.2 and even .sup.5D.sub.3 is occasionally observed, particularly in inorganic host lattices, which are associated with a much shorter decay time than the main .sup.5D.sub.0 emission. The features observed are consistent with such a mechanism, exhibiting a broad unstructured ligand-centered fluorescence within the first ca. 50 ns, followed by .sup.5D.sub.1.fwdarw..sup.7F.sub.J2 emission lines up to approximately 200 ns, and finally the long-lived .sup.5D.sub.0.fwdarw..sup.7F.sub.J2 lines which comprise the main emission overwhelmingly dominating steady-state spectra. In particular, it is noted that the .sup.5D.sub.1.fwdarw..sup.7F.sub.2 transition around 555 nm is a magnetic dipole transition, similar to the strongly dissymmetric .sup.5D.sub.0.fwdarw..sup.7F.sub.1 transition around 595 nm.

[0109] To find whether these short-lived features exhibit CPL activity, a CPL measurement is performed with 5 ns gate steps and width (FIG. 7C-E). For minimizing photoselection effects, a horizontal excitation polarization was used. Both ligand-centered (.sup.1LC) and .sup.5D.sub.1.fwdarw..sup.7F.sub.n transitions are clearly observed. The ligand-centered transition does not appear to be CPL-active; even if some dissymmetry were present, one may expect this to be far smaller than that of Eu.sup.3+ transitions and within the noise level of this measurement. In contrast, .sup.5D.sub.1.fwdarw..sup.7.sub.Fn transitions possess significant anisotropy, in particular the .sup.5D.sub.1.fwdarw..sup.7F.sub.2 transition at 555 nm has a bisignate CPL peak with |g.sub.lum|0.2.

[0110] For demonstration purposes, photoselection is intentionally induced by vertically polarized excitation and an S.sub.1 linear polarization measurement performed with 2 ns gate width and 0.5 ns gate steps, resulting in oversampled (partially overlapping) time bins (FIGS. 7F-H). As might be expected, some linear polarization is present in the .sup.1LC emission, which almost completely disappears within instrument response (around 2 ns, mainly limited by gate width). Though Eu.sup.3+ emission is observed, these features exhibit far less linear polarization, since these ionic transitions are from states populated only after ligand excitation and energy transfer. This implies that even when intentionally maximized, significant linear polarization is not present on timescales used for the CPL measurements in FIGS. 7C-E.

[0111] Nanosecond time resolution therefore allows to clearly resolve several transitions which were not observable before even in one of the most characterized CPL complexes, as they were drowned out by the long-lived main emission in steady-state and microsecond time-resolved measurements.

Chiral Organic Delayed Fluorescence Emitter

[0112] While strongly CPL-active chiral lanthanide complexes provide a convenient benchmark, studies on chiral emitters often involve materials with much weaker dissymmetry and faster luminescence. For example, purely organic small molecules in solution rarely exceed g.sub.lum=10.sup.310.sup.2 even in best-performing materials. Accordingly, a broader applicability with a CPL-active organic molecule is demonstrated. To introduce multiple timescales of emission, a recently developed chiral organic thermally activated delayed fluorescence (TADE) molecule is selected which has been previously characterized for CPL in the steady state. The molecular structure of this compound, here called (R/S)-BINOL-phthalonitrile-tBuCz, is shown in FIG. 8. Temporal characterization of CPL for this material, which combines ns-scale and s-scale decay with g.sub.lum of order 10.sup.3, is not feasible with pre-existing TRCPL W (methods.

[0113] FIGS. 8A-C show CPL spectra of R- and S-enantiomers (excitation 343 nm, 200 fs, 12.5 kHz). FIGS. 8D-F show time-resolved CPL spectra of the R-enantiomer. Prompt refers to the first 100 ns, Delayed to approximately 500 ns-80 s, and Total to a gate covering the complete emission process. FIG. 8G shows the total intensity decay kinetic (integrated over the full spectrum), with highlighted regions showing the Prompt and Delayed time regions and fitted constants for biexponential decay. FIG. 8H shows the dissymmetry factor (average from 445-455 nm) and total intensity decay (sum over 400-450 nm) kinetic as a function of time with 3 ns bins and 1 ns steps for the S-enantiomer (50 kHz repetition rate).

[0114] The two enantiomers show mirror-image steady-state CPL spectra (FIG. 8A-C) as expected, and the obtained g.sub.lum values of +1.810.sup.3 and 1.310.sup.3 at peak emission wavelength for the R and S enantiomers, respectively, match well the reported g.sub.lum value of 1.610.sup.3 for this material. To introduce time resolution, two experiments are performed on R-BINOL-phthalonitrile-tBuCz. First, the prompt is separately gated, delayed and total emission components (FIG. 8D-F). The total emission kinetic (FIG. 8G) clearly displays a biexponential emission process; the time ranges used for prompt and delayed emission are shown by the shaded regions, whereas the time gate for the total emission covers the entire emission timescale. Compared to the delayed component, prompt emission is approximately 1000 times shorter-lived, but over 10000 times intense. Accordingly, the total emission is dominated by the prompt component, such that the prompt and total emission spectra in FIG. 8D completely overlap. Somewhat unexpectedly, there appears to be a slight spectral shift between the prompt and delayed components; the emitting state in TADF is expected to be .sup.1CT for both cases, so the spectra should be identical. In principle, this could arise from delayed fluorescence overlapping phosphorescence from either the .sup.3CT state or a .sup.3LE localized triplet, or other long-lived species, although usually phosphorescence is not observed at room temperature for a purely organic molecule in solution at room temperature. Regardless of this slight spectral shift, the dissymmetry of the delayed component is not significantly altered, and the CPL/g.sub.lum spectra (FIG. 8E/F) are indistinguishable within noise for the different time bins. This confirms that not only is the instrument sensitive enough to accurately quantify g.sub.lum on the order of 10.sup.3, but it can also do so for a temporally separated emission component comprising only about 1% of the total emission intensity.

[0115] Finally, for demonstrating higher time resolution for even low CPL dissymmetry, CPL data has been collected with 3 ns bins and 1 ns steps (FIG. 8H). As the bin size exceeds the time steps, one gets a rise from instrument response for the first few bins. As convolution and division are not commutative, dissymmetry factors within instrument response should not be interpreted. Yet, after instrument response, g.sub.lum remains constant over the measured 15 ns time range where prompt decay is observed. This demonstrates that ns tracking of g.sub.lum on the order of 10.sup.3 is feasible with the setup and concludes the first transient CPL characterization of a chiral TADE compound.

Discerning Polarization Artefacts and Relaxation in an Achiral Dye

[0116] An achiral standard dye (rhodamine B in water, used also widely as fluorescent staining agent in biology) is measured to demonstrate i) low-noise zero baseline when photoselection is minimized, ii) presence & time-evolution of various apparent polarization components when photoselection is induced, and iii) full Stokes vector ns time evolution.

[0117] First, steady-state results are presented in a relatively low-viscosity aqueous solution with horizontal excitation polarization to minimize photoselection effects and vertical excitation polarization to intentionally maximize photoselection effects. FIGS. 9A-C show steady-state Stokes-vector measurements (excitation 515 nm, 200 fs, 50 kHz) in a low-viscosity and environment with horizontal and vertical excitation polarizations to inhibit and induce photoselection effects, respectively. Vertical excitation polarizations to solid correspond lines, horizontal excitation polarizations to dashed lines. It is pointed out that all spectra overlap in FIGS. 9A, and in 9B and 9C, the dashed lines are all flat around 0. FIGS. 9D-F show time resolved intensity differences (excitation 515 nm, 200 fs, 50 kHz) over the Stokes polarization basis (normalized to total intensity maximum) in a high-viscosity environment with vertical excitation polarization to induce photoselection. FIG. 9G shows the time evolution of all Stokes components (averaged over a 10 nm range around the emission peak) in low-viscosity and high-viscosity environments with vertical excitation polarization (2 ns time bins, 0.5 ns time steps), mapping distinctly different depolarization pathways over time. Due to the time bins exceeding the time steps, earliest bins are within instrument response.

[0118] The overall emission spectrum is not impacted by excitation polarization (FIG. 9A) and with horizontally polarized excitation we observe a flat background around 0 with noise at or below 10.sup.3 for all Stokes vector components, as expected for an achiral small molecule in solution (FIG. 9B/C, dashed lines). When vertically polarized excitation is used, nonzero values for all polarization components are observed (FIG. 9B/C, solid lines). Expectedly, the S.sub.1 component (horizontal-vertical linear polarization) shows the greatest response. Small but nonzero S.sub.2 values might stem from an excitation beam which is slightly tilted (so the excitation polarization is not perfectly vertical in the detection axis), while a nonzero S.sub.3 component for an achiral sample represents a CPL artifact. Such artifacts are well known to plague CPL measurements generally including the two-channel CCD approach and can be attributed to imperfections in the optical properties of components, such as residual birefringence and dichroism.

[0119] For the purposes of the time-resolved CPL measurement, it is desired to explicitly quantify how time-evolving real linear polarization impacts resultant CPL artifacts over time. This requires the presence of time-evolving linear polarization on sufficiently long timescales to be properly characterized by the instrument. In solution, molecules can rotate to reorient themselves, and therefore any preferential orientation of excited molecules arising from polarization of the excitation light is lost over time. While this is a straightforward way to achieve time-evolving linear polarization of emission, in many solvents the rotational relaxation timescale for fluorophores are faster than the 2 ns instrument response of the present example setup, but can be increased through solvent viscosity. As the emission properties of Rhodamine B are somewhat environment-dependent, rather than selecting completely different solvents of low and high viscosity it was decided to use water as a relatively low-viscosity environment and to increase the viscosity by adding sucrose. In a high-viscosity environment with vertically polarized excitation to induce photoselection, a very S.sub.1 large component is present immediately after excitation (FIG. 9D), with emission becoming depolarized over time. Smaller S.sub.2 and S.sub.3 components are also measured and depolarized over time (FIG. 9E, F), with their shape and magnitude (relative to S.sub.1) consistent with the steady-state measurements in a low-viscosity medium (FIG. 9B).

[0120] The time evolution of Stokes parameters around the emission maximum in low and high-viscosity solutions are plotted in FIG. 9G. For the low-viscosity solution, depolarization occurs almost completely within instrument response and the Stokes parameters are clustered around 0. For the high-viscosity solution emission remains partially polarized even after 10 ns, tracing an approximately straight line through the S.sub.1/S.sub.2/S.sub.3 polarization space. The S.sub.3 component, representing a CPL artifact, is therefore proportional to the real S.sub.1 component (though smaller in magnitude) as expected. These experiments showcase how broadband transient full-Stokes vector tracking enables to effectively discriminate between differently polarized excited state species, and how to distinguish them from artifacts, not captured in a CPL-only detection.

[0121] The present setup is therefore capable of time-resolved broadband full Stokes vector luminescence characterization with ns time resolution, ms range and unprecedented g.sub.lum sensitivity on the order of 10.sup.4. Broad applicability of the present setup across various timescales and degrees of CPL activity is demonstrated. For the CPL standard Eu[(+)-facam].sub.3, literature results could be reproduced in the steady state and for s-scale time-resolved measurements. By going into the nanosecond regime for the first time, it was possible to separate out and investigate the emission dissymmetry or ligand-centered and higher Eu.sup.3+ excited states which were hitherto drowned out by the main Eu.sup.3+ emission channel in steady-state and s-scale measurements. Moving beyond strong CPL emitters, a chiral TADF molecule was successfully measured with g.sub.lum on the order 10.sup.3, both by separating out the prompt (ns) and delayed (s) components and by ns-timescale step-by-step time-evolution. Notably, this first transient CPL characterization of any chiral TADF emitter shows that the dissymmetry of the singlet emission remains unchanged after inter-system crossing and reversal.

[0122] An achiral dye was employed to demonstrate a very low-noise baseline, while quantifying linear polarization responses and CPL artifacts in a geometry where excitation polarization induces photoselection of the molecules. By slowing down molecular reorientation and hence depolarization, sensitive transient broadband tracking of all Stokes-vector components was demonstrated on the ns scale, showcasing its power to discriminate between various components and CPL artifacts otherwise missed.