Photoluminescence Spectroscopy System  More Resources 

physics education essay research university report assessment newcastle

Photoluminescence Spectroscopy System Richard Laugesen 16 February 2002 1 2 INTRODUCTION...........................................................................................................................1 EQUIPMENT..................................................................................................................................1 2.1 2.2 2.3 SYSTEM CHARACTERISTICS........................................................................................................3 SOURCE MATCHING OPTICS........................................................................................................3 UV OPTICS .................................................................................................................................4 3 CALIBRATION..............................................................................................................................5 3.1 HOW TO CALIBRATE...................................................................................................................5 3.1.1 Wavelength Calibration ....................................................................................................5 3.1.2 Intensity Calibration .........................................................................................................6 3.2 STABILITY OF CALIBRATION ......................................................................................................7 4 USING THE SYSTEM ...................................................................................................................8 4.1 TURNING THE SYSTEM ON ..........................................................................................................8 4.2 TURNING THE SYSTEM OFF.........................................................................................................9 4.3 TURNING THE ACTICURE ON AND OFF........................................................................................9 4.4 RECORDING SPECTRA ................................................................................................................9 4.4.1 Inserting slides ................................................................................................................10 4.4.2 Using the Acticure...........................................................................................................10 4.4.3 How to record a spectra..................................................................................................10 4.4.4 File management.............................................................................................................11 4.4.5 Correcting the spectra ....................................................................................................12 5 POLYMER ....................................................................................................................................13 5.1 5.2 5.3 5.4 POLYMER SPECTRA ..................................................................................................................14 POLYMER DECAY ....................................................................................................................15 UV INTENSITY .........................................................................................................................17 ADDITIONAL MATERIALS ........................................................................................................17 6 BIBLIOGRAPHY .........................................................................................................................18 Photoluminescence Spectroscopy System 1 Introduction Photoluminescence (PL) is the light emitted by a material upon relaxation from an excited state. The spectral structure of this light can reveal information about the processes involved in the excitation, decay, and photophysics of the material. Conjugated polymers are easily placed into an excited state with UV light. Their PL spectra exhibits a broad structure (~200 nm wide) located in the visible range and may contain multiple peaks. One peak is due to the electronic ground state of the polymer (generally nondegenerate) the others are vibronic states. The broadening of the peaks is due to the range of energies each states may have. A system for making routine PL spectroscopy measurements of polymer slides was constructed. 2 Equipment A Spex 1681C Minimate 0.22 m spectrometer (f/4 aperture) with a Princeton Instruments IRY detector head was used to record PL spectra. A computer (486SX) was used to interface with the detector. The detector requires nitrogen and a water cooling system (NESLAB Endocal RTE-5DD). The detector head was a proximity focused microchannel plate intensified diode array (IRY1024G/RB). It is gateable from 5 ns to 6 ms, has a quartz window and a useable wavelength of 170 nm to 920 nm. The detector head was controlled by an optical spectrometric multichannel analyser (OSMA) which was in turn controlled by the computer. The dial on the front of the spectrometer is inaccurate and should not be used other than for rough observations. A proper wavelength calibration should be performed as described in Section 3.1.1. It only reads the correct central wavelength when a 1200 g/mm grating is used, for an X g/mm grating the correct wavelength is; λreal = λdial ⋅ 1200 X An EFOS Acticure high pressure mercury short arc lamp was used to excite the samples, a filter which gave a peak wavelength of 365 nm was used. A Kodak 18A filter was used to remove the visible component of the UV source. An Eletro-lite ELS-300 UV Radiometer was used to measure the light source power. Glass lenses were used for focusing, other miscellaneous components were used for mounting. Light shielding involved two solid screens and a double layer of black cloth. With the lights directly above the spectrometer turned off this is sufficient as can be seen in Figure 1. 1 Photoluminescence Spectroscopy System Figure 1. Spectra of background light with black sheets covering equipment. Notice that the mercury fluorescence lights are clearly evident when the lights are turned on. Although not very intense, polymer PL spectra have peak intensity from 0.1 up to 6.5 (a.u.). See Figure 2 for a photograph of the experimental apparatus. Figure 2. Photograph of PL system 2 Photoluminescence Spectroscopy System 2.1 System characteristics The spectrometer was found to have a resolution of 0.96 nm by measuring the width at half-maximum of the green mercury 546.074 nm line (isolated with an interference filter). The natural line width of this line is ~0.001 nm, therefore the line broadening must be due to the spectrometer. A grating with 300 g/mm was used for the PL measurements as it gives a large wavelength window; ~240nm. Other gratings are available, 1200 g/mm and 2400 g/mm, but have wavelength windows too small for the broad spectral structure of the polymers to be seen in their entirety. The UV Acticure has a power range of 33.9 mW/cm2 to 3.36 W/cm2 and was found to be very stable. 2.2 Source matching optics Source matching optics are required to optimise the light gathering ability of the spectrometer and hence improve the resolution. If the source is too close to the spectrometer slit then the incident mirror will be over-filled and stray light will enter the spectrometer cavity degrading the signal. If the source is to far from the slit then the mirror will be under-filled and the light gathering power of spectrometer will not be used to its full potential. Source matching optics solves the problem of finding an optimum distance for the source. Lens 1 is matched to the source to create parallel. The focal length and distance of lens 2 is chosen to match the cone of acceptance of the spectrometer. The distance of lens 1 from lens 2 is not important, the space being used to insert neutral density filters. See Figure 3. Beam dump Filters Incident Lens 1 Lens 2 UV optics Slide in mount Spectrograph Figure 3. System setup. Dashed lines are PL light, Solid line (orange) is UV light. 3 Photoluminescence Spectroscopy System Figure 4. Photograph of optics. Lens 1 had a focal length of 150 mm, was placed 170 mm from the sample holder, and was 230 mm from lens 2. Lens 2 had a focal length of ?? mm and was placed at 90 mm from the slit 1. 2.3 UV optics The excitation light from the UV source must be focused onto the slide. This was accomplished by aperture stopping the output of the Acticure, allowing a relatively collimated beam to pass through lens 3 which focused it to a small dot on the slide, see Figure 5. The size of this dot may be adjusted by the aperture size; from 3 mm in diameter to covering the entire slide. Lens 3 had a focal length of 50 mm and was approximately 100 mm from the slide. A calculation2 of the power at the slide was made by measuring the total power at the Acticure output using the radiometer. It was found that 1.0 W/cm2 at the Acticure output resulted in 1.41 mW/cm2 at the slide (4 mm beam width at the slide). This is too weak for non-linear effects. 1 2 It is from the fibre optic adaptor. This calculation is an approximation as the collimation of the UV beam from the output was ignored. 4 Photoluminescence Spectroscopy System Aperture Slide in mount Acticure Output Lens 3 Figure 5. UV optics. 3 Calibration Calibration involves two steps; wavelength calibration, and intensity calibration. The first and last 50 diodes are ignored as they give low response. Wavelength calibration involves converting the diode numbers to corresponding wavelength values. Intensity calibration is necessary to correct for the effect the spectrometer, detector, and other optics has on the observed spectra. If the spectra is simply taken as is from the spectrometer it is a convolution of the true spectra of the sample, and the spectral response of the grating, the slit, the diodes comprising the detector array (S20), neutral density filters, all other optical components in the system; including external to the spectrometer, and the dark charge spectrum of the detector. All these effects must be removed from the recorded spectra to observe the true spectra of the sample. 3.1 How to calibrate 3.1.1 Wavelength Calibration Wavelength calibration involves matching known atomic electronic transition lines to the diode number they are observed at. When many pairs of diode-to-wavelength values have been collected, a 2nd order polynomial regression fit3 is performed. The wavelength corresponding to every diode in the array may then be calculated from this polynomial. Found that the easiest way to match observed lines with reference lines (ie NIST – http://physics.nist.gov/cgi-bin/AtData/lines_form) was to do a rough fit with 3 or 4 strong lines first, then use this fit to get approximate wavelength values for the others. Take care not to saturate the detector when using atomic discharge sources, neutral density filters will be required. The source should be placed at the same location as the slide holder to maximise resolution. An example polynomial fit for 23 mercury/neon/sodium lines; λ = (−2.361 ± 0.788) −6 d 2 + (0.243 ± 0.001)d + (463.422 ± 0.279) ; r 2 = 1 A linear fit of the same lines; 3 The linest function in excel was used for the regression. 5 Photoluminescence Spectroscopy System λ = (0.240 ± 0.0003)d + (464.121 ± 0.180) ; r 2 = 1 Where λ is the corresponding wavelength for the diode d. The average squared difference between the calculated wavelength and the measured wavelength (of the lines used in the calibration) was 0.19 nm for the polynomial fit and 0.23 nm for the linear fit. There is spread of 0.24 nm over each diode which limits the precision with which a line may be located onto a single diode, this seems to be reflected in the average difference between measured and calculated line positions and can be considered as the uncertainty in wavelength. Although the detector response is quite linear a polynomial fit is still preferred as it gives a lower average error. It should be noted however that if a limited number of lines (<8) are used for the regression then the polynomial fit may give wavelength values which drastically change with only slight changes in diode number. 3.1.2 Intensity Calibration To compensate for the dark charge spectra of the detector, switch the detector to gate (on the rear of the detector head) and record a spectra with a large number of accumulations, see Section 4.4.3 concerning recording spectra. Subtracting this spectra from all subsequent spectra recorded with the system is necessary and will remove the effect of the dark charge. The exposure time is critical, dark charge spectra recorded at a particular exposure time may not be applied to sample spectra taken with a different exposure time. The simplest way to remove the other unwanted spectral effects is the following. Record the spectra of a radiometrically calibrated light source; essentially this is a light source which we know the spectra of. Tungsten and deuterium calibrated light sources are available from a few sources but are expensive. Divide the known calibration data of the source spectra by the recorded spectra (dark subtracted & normalized) point-for-point, the resulting factors contain the unwanted spectral information for each diode; correction factors. Multiplying the spectra of an unknown sample (dark subtracted) by these correction factors point-for-point will leave its true spectra. Since a certified radiometrically calibrated light source was not available a Solux halogen lamp was used, spectral calibration data (80 points) was obtained from the company; Eiko. This data was smoothed and interpolated for the region of our wavelength window with 924 data points; the number of diodes used. Since the halogen lamp is such an intense light source extra care was taken not to saturate the spectrometer. The lamp had an aperture placed directly in front of it, and ground glass mounted in place of the slide holder to produce diffuse light replicating the quality of the PL, neutral density filters were also required to reduce the intensity. The intensity calibration has not been verified using a standard fluorescence source4. Since a noncertified light source was used for calibration and the conditions of its spectral reference data are 4 A sample of the fluorescence standard, quinine sulfate dihydrate has been ordered. 6 Photoluminescence Spectroscopy System unknown it is critical that a verification is performed. However comparisons of recorded polymer PL spectra are in good agreement with those in the literature, see Figure 6. Figure 6. Normalised PL spectra of the polymer MEH-PPV (ours has 1% nanotubes). Notice the agreement in peak positions. JAK spectra is from [1]. An analytical process of deconvolutions could be attempted to remove the unwanted spectral effects, but the approach stated above is the simplest as it includes any unthought of elements that may effect the spectra are automatically included in the correction factors. The interpolation used for the intensity calibration gives a linear wavelength spread, our detector response is marginally non-linear, and therefore the interpolation and the wavelength calibration values will be slightly different and may have an effect on the correction factors. The difference of the centre diode wavelength is ~0.001 nm, significantly smaller than our uncertainty in wavelength of 0.24 nm and may be ignored. 3.2 Stability of calibration The calibration was found to be relatively robust. The source matching optics were disturbed (included replacing lenses) and reassembled, the correction factors were recalculated but no difference was found between these factors and those before the disturbance. Over a 21 day period of normal laboratory use a drift in wavelength of 0.5 nm was observed. This suggests a weekly recalibration would be sufficient. The system will obviously need to be recalibrated if the grating is rotated or replaced. The dark charge spectra was recorded over a 6 hour period to investigate the temperature stability of the detector. A variation of only 0.4% in average intensity was observed and equilibrium was reached after 4.5 hours. The spectral shape did not change, see Figure 7. 7 Photoluminescence Spectroscopy System Figure 7. Each data point is the average of the dark charge spectra of the detector head. This is a good measure of the temperature stability of the system. Notice only a small variation; ~0.4% and equilibrium after 5 hrs. 4 Using the System Care must be taken when using the equipment as the detector head may be damaged easily. If the detector control unit beeps, it means something is wrong; usually too much light is entering the spectrometer and saturation has occurred. Quickly cover the entrance slit to the spectrometer and take measures to reduce the level of light entering the slit. If the unit continues to beep when covered then switch off the detector controller immediately and consult the manual. 4.1 Turning the system on This following procedure must be followed when turning the system on to avoid damaging the detector head; 1. 2. Check that the water level of the cooling unit is ~3 cm from the top. Switch on the water cooling unit. Check that the set-temperature is 18.1oC by holding down the set button. Caution: Have found that the set-temperature may change on its own accord, suggest that it is monitored regularly while making measurements. If the temperature rises too high the head may be damaged, if it drops below the dew point of the air, condensation will form on the hoses. 3. Check the nitrogen pressure in the cylinder. Turn the regulator on (~100 kPa) and adjust the flow rate of nitrogen with the valve on the nitrogen hose near the spectrometer. The flow rate is correct when one can not feel the flow with fingers but can on moistened lips. Firmly push the nitrogen hose into the detector head. 8 Photoluminescence Spectroscopy System 4. 5. Wait for the temperature of cooling unit to reach ~18.1oC. Switch the detector controller unit on. The LED will be red, wait for it to turn green. If the unit beeps something is wrong, turn it off immediately. Once the LED is green the system is ready for recording spectra. 6. Switch the computer on. The manual recommends waiting at least half an hour before taking any measurements. However the variation in detector response is small (see Figure 7) and suggests that measurements could be made as soon as the LED turns green. 4.2 1. 2. 3. 4. Turning the system off Switch off the computer. Switch off the detector controller head. Unplug the nitrogen hose and turn off the regulator. Switch off the cooling unit. 4.3 Turning the Acticure on and off To turn the unit on follow this procedure; 1. 2. Turn unit on with the switch at the back. Turn the lamp on by pressing the lamp power button (there is a slight delay). The display will flash while the lamp prepares itself. If the display reads cool wait a few minutes while the lamp cools before trying again. 3. When the display stops flashing the unit is ready to use. To turn the unit off follow this procedure; 1. 2. Turn off the lamp by pressing the lamp power button. Wait approximately 5 minutes for the lamp to cool before turning the unit off with the switch at the back of the unit. 4.4 Recording Spectra If program does not automatically start when computer is turned on change to the directory c:/st120/ and type nsma. The program will display a menu with nine options. 9 Photoluminescence Spectroscopy System 4.4.1 Inserting slides Insert the slide into the slide holder, take care not to get fingerprints on it as they fluoresce. Neutral density filters may be required depending on the intensity of the UV source and the properties of the material. 4.4.2 Using the Acticure Pressing the II button (hold down for a couple of seconds) allows the lamp output to be left on continuously and the intensity adjusted using the arrow keys either side of this button. Monitoring of the intensity is done via the radiometer. To turn the lamp output off press the lock/unlock button. The start/stop button may be used if a specific length of UV exposure is required. The choice of UV intensity is a trade off between the rate of the decay of the polymer and the increased signal-to-noise ratio which comes with greater intensity, see Section 5.2 for detailed look at MEH-PPV decay under different intensities. Greater intensity will also result in more UV scatter in the system which may enter the spectrometer. Although the wavelength window may not be in the UV, the diodes comprising the detector array are sensitive in UV and therefore the stray light may cause a saturation of the detector. An intensity of 1.4 mW/cm2 (1.0 W/cm2 at the Acticure output) has been used for most measurements. 4.4.3 How to record a spectra 1. Select option 1, “Initialise Experiment”, this displays another menu where the experimental parameters may be set up. 2. Select option 2, “Experiment Name”, and enter a descriptive title for the file (without extension), this filename must have 8 or less characters. 3. Select option 7, “Number of Accumulations” to enter the number of spectra to be accumulated that will contribute to the final spectra, this improves the signal to noise ratio. A sufficient value is 5005, if too large a value is entered then the accumulated intensity may max out the program (result is negative numbers). Increasing the number of accumulations improves the signal-to-noise ratio according to a 4. 5. N relationship, where N is the number of accumulations. Press enter to return to the main menu. Select option 2, “Run Experiment” to begin recording a spectra6. The spectra will be displayed on the screen in real-time and stored automatically, F9 toggles storing of spectra. The horizontal axis is diode number, the vertical axis is intensity in arbitrary units. If a number of accumulations has 5 6 With an exposure length of 0.033 s, this equates to an 16.5 s integration time, The first spectra recorded when the computer is turned on will fail due to a “data underrun” error, just ignore and try again. 10 Photoluminescence Spectroscopy System been selected the spectra will appear to progressively grow as each spectra is accumulated until the number of accumulations selected is reached, at which point it will start accumulating again. Pressing F10 auto-resizes the screen to fit the spectra. F5 and F6 decrease and increase the intensity scale respectively. F3 and F4 decrease and increase the diode number scale respectively. 6. Press escape when you have recorded enough spectra for your purposes. The file is stored with the extension .spe in the c:/st120 directory. The .spe files have a file size of approximately 8.2 kb per spectra. Other options commonly changed; 1. “Scan units per exposure”; changes the exposure time used to record each spectra. Note that this will require an appropriate dark charge spectra for the calibration process. Generally left on 1 sample per exposure, 0.033 s exposure time. 2. “Store 1 out of N spectra”. Useful for collecting spectra at regular but large intervals of time. For example to store a spectra every 5 minutes, every 18th spectra should be stored (500 accumulations and 1 sample per exposure). Saves on hard disk space. Note that when this feature is used the number of stored spectra indicator in the top left of the spectrum screen indicates all spectra not the number of stored spectra. 4.4.4 File management The file format that is saved includes data specific for this program and is not needed. Therefore an ASCII file is preferred. 4.4.4.1 1. 2. 3. 4. Making an ASCII file From the main menu select option 6, “Convert experiment to floating point/ASCII format”. Type the filename wanting to convert (no extension). Type a name for the ASCII file, generally the same name is used. First spectra, number of spectra, and step size apply to the range of spectra from the file wanted in the ASCII output. The file will be saved with the extension .prn in the c:\st120 directory. The file size depends on the number of spectra, approximately 21.5 kb per spectra. 4.4.4.2 Transferring via disk To transfer the file from the equipment computer to another computer the file must be transferred via floppy disk. 11 Photoluminescence Spectroscopy System 1. 2. From the main menu select option 9, “Execute a DOS program”. Insert a floppy disk in drive with adequate space. (to check the size of the file type dir filename.prn then return to step 1) 3. 4. Type copy filename.prn a: Take this disk to another computer. 4.4.5 Correcting the spectra The spectra file (.prn) must have diode number converted to wavelength and be corrected for intensity variations introduced by the equipment. A program was custom written called spec.exe which performs these operations. 4.4.5.1 Using spec.exe spec.exe uses a command line interface; spec [filename.prn] [number of accumulations] [number of spectra] Where [filename.prn] is the output of the nsma program, [number of accumulations] is the number of accumulations used when recording the spectra, and [number of spectra] is the number of spectra to correct from the .prn file and save in the resultant output files. The file spec.cfg is required, this contains information required for the correction. See below for instructions on generating or modifying this file. The output of spec.exe is two ASCII files with the same filename as the .prn file, but with the extensions .spx and .int. The .spx file contains the corrected spectra and has the format; [wavelength] [spectra 1] [spectra 2] [spectra 3] . . . – for each diode Where [spectra 1], [spectra 2], and [spectra 3] are the corrected intensity values for each spectra. The .int file contains the integration of each spectra. It has the format; [spectra number] [integration value] – for each spectra The .spx and .int files may now be imported into another program for analysis. 12 Photoluminescence Spectroscopy System 4.4.5.2 Generating spec.cfg spec.cfg is required for spec.exe and contains all the calibration details required to correct the spectra. Whenever the system requires recalibration this file must be updated. The format of spec.cfg is; [wavelength/correction factor filename] [dark charge filename] [number of accumulations for dark charge] The wavelength/correction factor file is generated using another method (ie. Excel). The method of determining the wavelength value and correction factor for each diode is described in Section 3. Its format is; [wavelength] [normalized correction factor] - for each diode The dark charge file is a dark charge spectra (.prn file from nsma program) recorded as described in the Section 3.1.2. 5 Polymer Comparing the PL spectra of polymer slides is difficult because of processes affecting the reproducibility of the measurements. As well as the total intensity of PL light increasing or decreasing depending on the thickness of the polymer, the spectral shape may also change. The level of order within the polymer has an effect on the spectra; the greater the order the more defined the peaks are, a slide may even have local regions of high order and regions of low order. Controlling the order on the slide is beyond our control. Varying thickness of the polymer across the slide will also contribute to local effects. Therefore the region that the UV beam excites will determine the structure observed in the spectra, see Figure 8, and two slides made of the same polymer and under the same conditions may not give the same spectra. 13 Photoluminescence Spectroscopy System Figure 8. Normalised PL spectra from exciting different regions of the same PPV/ITO slide. 5.1 Polymer spectra The spectra of a number of polymer were recorded, all were recorded under the same conditions at a UV intensity of ~1.4 mW/cm2. They were smoothed with a 20 point adjacent averaging function. 14 Photoluminescence Spectroscopy System 5.2 Polymer Decay Investigated the decay of MEH-PPV during UV exposure (1.4 mW/cm2), see Figure 9. The integrated PL output of the polymer decreases over time as the UV causes photo-oxidation. A 1st and 2nd order exponential decay was fitted to the data. The 1st order fit7 gave a time constant of 4 min 24 s but does not appear to fit the data very well. The 2nd order8 fits the data much better and gives time constants of 6 min 41 s and 36 s. This suggests that there are 2 processes involved in the decay, interestingly the MEH spectra has 2 peaks, which may be decaying at different rates. 7 I = (0.49) + (0.44) ⋅ e I = (0.45) + (0.42) ⋅ e − t ( 4.41) t ( 6.69 ) − t ( 0.60 ) 8 − + (0.13) ⋅ e 15 Photoluminescence Spectroscopy System Figure 9. Decay of MEH-PPV during UV exposure. Solid line is a 1st order exponential decay fit, dashed line is a 2nd order fit. The shape of the spectra does change as the polymer decays, see Figure 10. The spectra has shifted slightly to the blue and the 2nd shoulder has decreased in size (relative to the 1st). This change in spectral shape can not be due to local effects as the region being excited did not change (slide was not removed & reinserted). Figure 10. Normalised PL spectra of fresh MEH-PPV and same slide after 14 min of UV exposure. 16 Photoluminescence Spectroscopy System 5.3 UV intensity A higher UV intensity causes a larger PL intensity. Slight changes are seen in the spectral shape of MEH-PPV, no change is seen in PPV, see Figure 11. This could be because the MEH-PPV slide had already decayed significantly (10 min of UV exposure) when the higher intensity spectra was recorded, whereas PPV does not decay as quickly. The two PPV spectra are almost identical point-for-point. Therefore ignoring non-linear effects that may occur at extremely high intensities we may say that the spectral shape does not depend on intensity. Figure 11. Effect of excitation intensity on the PL spectra. (actually a PPV/ITO slide not PPV) 5.4 Additional Materials The spectra of slides with an additional layer of material were recorded. The additional materials have caused the spectra of PPV to change shape; 3 peaks remain, however their relative heights have changed. The PSS layer has even caused a red shift; 7 nm for the 1st peak and 4.8 nm for the 2nd peak. A detailed analysis is beyond the scope of this document. 17 Photoluminescence Spectroscopy System The PL spectra of a blend of PPV with nanotubes was recorded. The nanotubes have caused a decrease in overall PL intensity, a change in relative peak heights, and a slight red shift of 1.2 nm for the 1st peak and 4.1 nm for the 2nd peak. This shift is similar to that seen in the PPV/PSS, where the 1st peak was also shifted by more than the 2nd. 6 Bibliography [1] Rachel Jakubiak, Christopher J. Collision, Wai Chou Wan, and Lewis J. Rothberg, "Aggregation Quenching of Luminescence in Electroluminescent Conjugated Polymers", J. Phys. Chem. A 1999, 103, 2394-2398. [2] User manuals for equipment, especially the hardware manual for the detector head. 18