A Technical Guide for Performing Spectroscopic Measurements on Metal-Organic Frameworks

Summary

Here, we use a polymer stabilizer to prepare metal-organic framework (MOF) suspensions that exhibit markedly decreased scatter in their ground-state and transient absorption spectra. With these MOF suspensions, the protocol provides various guidelines to characterize the MOFs spectroscopically to yield interpretable data.

Abstract

Metal-organic frameworks (MOFs) offer a unique platform to understand light-driven processes in solid-state materials, given their high structural tunability. However, the progression of MOF-based photochemistry has been hindered by the difficulty in spectrally characterizing these materials. Given that MOFs are typically larger than 100 nm in size, they are prone to excessive light scatter, thereby rendering data from valuable analytical tools like transient absorption and emission spectroscopy nearly uninterpretable. To gain meaningful insights of MOF-based photo-chemical and physical processes, special consideration must be taken toward properly preparing MOFs for spectroscopic measurements, as well as the experimental setups that garner higher quality data. With these considerations in mind, the present guide provides a general approach and set of guidelines for the spectroscopic investigation of MOFs. The guide addresses the following key topics: (1) sample preparation methods, (2) spectroscopic techniques/measurements with MOFs, (3) experimental setups, (3) control experiments, and (4) post-run stability characterization. With appropriate sample preparation and experimental approaches, pioneering advancements toward the fundamental understanding of light-MOF interactions are significantly more attainable.

Introduction

Metal-organic frameworks (MOFs) are composed of metal oxide nodes linked by organic molecules, that form hierarchical porous structures when their constituent parts react together under solvothermal conditions¹. Permanently porous MOFs were first reported in the early 2000s, and since then, the burgeoning field has expanded to encompass a wide range of applications, given the unique tunability of their structural components^2,3,4,5,6,7. During the growth of the field of MOFs, there have been a handful of researchers that have incorporated photoactive materials into the nodes, ligands, and pores of MOFs to harness their potential in light-driven processes, like photocatalysis^8,9,10,11, upconversion^{12,13,14,15,16}, and photoelectrochemistry^17,18. A handful of the light-driven processes of MOFs revolve around energy and electron transfer between donors and acceptors^{17,19,20,21,22,23,24,25}. The two most common techniques used to study energy and electron transfer in molecular systems are emission and transient absorption spectroscopy^26,27.

A great deal of research on MOFs has focused on emission characterization, given the relative ease in preparing samples, performing measurements, and (relatively) straightforward analysis^{19,22,23,24,28}. Energy transfer typically manifests itself as a loss in the donor emission intensity and lifetime and an increase in the emission intensity of the acceptor loaded into the MOF backbone^19,23,28. Evidence of charge transfer in a MOF manifests as a decrease in emission quantum yield and lifetime of the chromophore in the MOF^29,30. While emission spectroscopy is a powerful tool in the analysis of MOFs, it only addresses part of the necessary information to present a complete mechanistic understanding of MOF photochemistry. Transient absorption spectroscopy can not only provide support for the existence of energy and charge transfer, but the method can also detect spectral signatures associated with the non-emissive singlet and triplet excited state behaviors, making it one of the most versatile tools for characterization^31,32,33.

The primary reason why more robust characterization techniques like transient absorption spectroscopy are seldom applied toward MOFs is due to the difficulty in preparing samples with minimal scatter, especially with suspensions³⁴. In the few studies successfully performing transient absorption on MOFs, the MOFs are <500 nm in size, with some exceptions, stressing the importance of reducing particle size to minimize scatter^{15,21,25,35,36,37}. Other studies make use of MOF thin films¹⁷ or SURMOFs^38,39,40 to circumvent the scatter issue; however, from an applicability stand-point, their use is quite limited. Additionally, some research groups have taken to making polymer films of MOFs with Nafion or polystyrene³⁴, the former raising some concerns for stability given the highly acidic sulfonate groups on Nafion. Gaining inspiration from the preparation of colloidal semiconductor suspensions^41,42, we have found great success using polymers to help suspend and stabilize MOF particles for spectroscopic measurements¹¹. In this work, we establish widely applicable guidelines to follow when it comes to preparing MOF suspensions and characterizing them with emission, nanosecond (ns), and ultrafast (uf) transient absorption (TA) spectroscopy techniques.

Protocol

1. Preparation of MOF suspensions using a polymer stabilizer

1. Weigh out 50 mg of bis-amino-terminated polyethylene glycol (PNH₂, M_n ~1,500) (see Table of Materials) and transfer to a one-dram vial (Table of Materials). Weigh out 1-5 mg of PCN-222(fb) (see synthetic protocol¹¹) and place it into the same vial with PNH₂.
   NOTE: To attain the best possible MOF suspensions, the synthetic conditions required to make the MOF particle sizes need to be at or below 1 µm.
2. Find a suitable solvent (if not water, use an anhydrous solvent such as dimethylformamide [DMF] or Acetonitrile [ACN]; see Table of Materials) to suspend the MOF, making sure that the solvent window is wide enough so that the solvent itself is not excited with the wavelength of choice. Transfer 1-3 mL of solvent to the vial using an autopipette fitted with an appropriate pipette tip.
   NOTE: The commonly used solvents noted above have wide solvent windows-CH₃CN: 200 nm high-energy cut-off; and DMF: 270 nm cut-off. Solvents with higher refractive indices (1.4-1.5), like DMF, DMSO, and toluene, can be used to help minimize light scatter by being more closely aligned with the refractive index of quartz glass (ca. 1.46-1.55), thereby minimizing the bending of light in unwanted directions when passing through the cuvette.
3. Using a tip sonicator (see Table of Materials), sonicate the vial contents for 2-5 min at 20%-30% amplitude (i.e., the distance in which the sonicator tip longitudinally moves, typically ~60 µm for a 3 mm diameter probe at 30% amplitude) with intervals of 2 s on and 2 s off. This procedure serves to break up MOF aggregates and helps coat MOF particles with polymer. Ensure that the MOF suspension is well-dispersed and homogeneous by the end of the sonication process.
   NOTE: Sonication times vary depending on how well the MOF inherently disperses.
4. Open up a fresh 10 mL plastic syringe (Table of Materials) and draw up the suspension into the syringe. Remove the syringe needle and replace it with a polytetrafluoroethylene (PTFE)-mesh 200 nm syringe filter (Table of Materials). Pass the suspension through the syringe filter into a new clean vial. The resulting suspension is ready for transient absorption spectroscopic measurements.
   ​NOTE: Given that the average particle sizes of some MOFs exceed 200 nm, choosing the appropriate size is up to the user's discretion.

2. Preparation of filtered MOF suspensions for nanosecond transient absorption measurements (nsTA)

1. With the filtered MOF suspension obtained in section 1, the absorption spectrum of the suspension needs to be obtained (Table of Materials). Wash a cuvette (1 cm path length) that is capable of sealing and purging (Table of Materials) with solvent three to five times, then fill it with 3 mL of DMF.
2. With an absorption spectrophotometer, choose a wavelength region to measure the solvent and suspension. Measure the blank ultraviolet-visible (UV-Vis) spectrum of the solvent in the cuvette and set it as a background scan to be subtracted from the sample scans. Empty the solvent contents of the cuvette and transfer the MOF suspension into the cuvette.
   NOTE: The electronic absorption spectrum of the PNH₂ stabilizer ( ~250 nm) has a weak absorption tail that persists until 450 nm, with an absorbance of ~0.01 at 450 nm at the initial concentration.
3. Measure the absorption spectrum of the initial MOF suspension, keeping in mind the desired wavelength needed to excite the MOF sample. If the MOF suspension has an absorbance or optical density (OD) >1 at the desired excitation wavelength, then dilute with solvent and measure the absorbance spectrum until the OD is ≤1 at the excitation wavelength.
   ​NOTE: For narrow-angle transient absorption measurements, repeat the absorption measurements to attain an appropriate OD at the excitation wavelength using a 2 mm cell (Table of Materials). For nanosecond transient absorption measurements of MOFs, an absorbance or OD of 0.1-1 at the excitation wavelength is needed to follow Beer's Law. The required OD is a wide range because some samples absorb strongly in different regions. A perfect example of this is porphyrins. Porphyrins have a strong narrow Soret band transition between 400-450 nm, whereas their Q-band transitions between 500-800 nm are quite weak. If one wanted to excite at one of their Q-bands, prepping a solution with an OD ~0.5 at one of the Q-bands would consequently exhibit a Soret band absorption >3, and the transient absorption detector would not be able to quantitatively process the changes in this region. Ultimately, it is at the user's discretion to determine the appropriate excitation wavelength and absorbance amplitude that allows for quantitative measurements in the desired spectral window.

3. Purging the MOF suspension

1. With the MOF suspension adjusted to have the desired absorbance spectrum for TA measurements, place a 2 mm x 8 mm stir bar (Table of Materials) into the cuvette, and seal the inlet cuvette joint with a rubber septum.
2. Take a 1 mL plastic syringe (Table of Materials), cut half of it with a pair of scissors, and keep the half of the syringe that allows for needles to be attached to it.
3. With one end of a flexible tube attached to an Ar or N₂ tank (Table of Materials), insert the syringe half into the other end of the tubing with the needle end outside.
4. Wrap the stem of the exposed syringe half with parafilm (Table of Materials) to create a seal with the tube and syringe. If a hose clamp is available, it can be used instead of parafilm to create a seal with the syringe and purging tube.
5. Attach a long purge needle (3 in, 25 G) (Table of Materials) to the syringe end and insert the needle into the sealed cuvette into the suspension. Take the needle from the 1 mL syringe (step 3.2) and insert it into the cuvette. Turn on the Ar or N₂ flow and purge the suspension for 45 min-1 h.
   ​NOTE: A technique known as "double-purging" is often used for solvents with a boiling point <100 °C. To employ this technique, a sealed flask with solvent is purged with the inlet needle, with one end of a cannula inserted into the flask headspace and the other end of the cannula inserted into the cuvette suspension. The outlet needle is inserted into the cuvette headspace. Purging this way minimizes solvent loss from evaporation over time.
6. After purging is finished, remove the needles and wrap the cuvette septum with four to five 2 in slices of parafilm (Table of Materials). Measure the absorption spectrum of the sample to make sure it matches the standards set in section 2. The sample is now ready for transient absorption measurements.

4. Perpendicular pump-probe nanosecond transient absorption setup (nsTA)

1. Turn on the laser and nsTA spectrometer systems (Table of Materials; Figure 1). Adjust the laser output power to a low enough level, such that placing a white business card in the beam path allows for clear visibility of the laser spot, but not so bright that it is blinding.
2. Open up both the mechanical laser shutter and the probe beam shutter so that they are both in the path of the sample holder.
3. Adjust the vertical and horizontal position of the laser beam (Figure 1, P3), such that it hits the center of the sample cell holder (Figure 1, SC1) where the sample will be placed. Use a business card to confirm the position. Place a neutral density (ND) filter (OD 2; Table of Materials) in the path of the probe beam.
   NOTE: All mirrors and prisms present in the nsTA system described herein are mounted on kinematic mounts (Table of Materials), and the beam positions are adjusted by manually turning the vertical and horizontal knobs on the mounts. A longer, 1 cm wide white card can be placed across the inside of an empty cuvette in SC1 to make for an easier alignment.
4. Place a cut business card (~1.5 cm width) in the sample holder (or hold it in the sample holder) and angle it across the sample holder such that both the laser beam and probe beam are hitting the same side of the business card. Fine-tune the laser beam position vertically (P3) to obtain the best overlap with the most intense part of the probe beam.
5. Close the shutters, remove the ND filter, and place the sample in the sample chamber along with a magnetic stirrer (Table of Materials). TA measurements can now be taken.
6. The system and software used in this work are provided in the Table of Materials. In the software suite, there are selection boxes titled Spectral Absorption and Kinetic Absorption for measuring TA spectra and absorption kinetics, respectively. Select the Spectral Absorption button and select the Setup mode.
7. Set the time zero of the laser pulse in the software setup window by adjusting the input time in increments of +0.010 µs (e.g., -0.020 µs, -0.010 µs, 0.000 µs, 0.010 µs, etc.) until the laser pulse is not observed in the probe beam spectrum.
8. With the time zero set, optimize the amount of light hitting the charge-coupled device (CCD) detector in the setup window by adjusting the bandwidth, gain, and gate width to reach a high enough count to obtain an adequate signal while not saturating the detector.
   NOTE: We leave this process up to the user as detectors vary from system to system.
9. To collect a TA spectrum, click on the Multiple button in the Spectral Absorption tab. Make sure the settings from the setup window are present in this window. If the sample emits light, click on the Backgrounds tab and click on the Subtract Fluorescence Background button. For a cursory scan, set the averages to 4 to make sure a quality initial TA spectrum is obtained. If a satisfactory TA spectrum is obtained, obtain another one with more averages.
10. To map a TA spectrum at different time delays after time zero, select the Map button in the Spectral Absorption tab. Ensure that the setup parameters have not changed in this tab. Enter the desired time intervals for mapping, click on Apply, and then click on Start to map the spectra.
11. To obtain absorption kinetics at desired wavelengths, click on the Kinetic Absorption button in the software and click the Setup button in the drop-down menu. Enter the wavelength of interest in the Controller tab in the setup window and adjust the bandwidth to a suitable level. Usually, a bandwidth of 1 nm is a good starting point.
12. In the Oscilloscope tab, adjust the photomultiplier tube (PMT) detector time window so that it is long enough to see the entire kinetic trace, from before laser excitation to the signal fully decaying back down to baseline. A usual starting point is a 4,000 ns window. Adjust the PMT voltage range to a suitable level, where the entire TA trace is observable on the signal axis. A voltage range of 160 mV is reasonable to begin measurements. Click on Apply and then Start. If the signal is too low or the time window is too short or too long, click on Stop and adjust the bandwidth and time window to suitable levels, making sure not to set the bandwidth too high to saturate/damage the detector.
13. With the kinetic trace properly set up, close the Setup window and open the Multiple window from the drop-down menu after clicking on the Kinetic Absorption button. Check to make sure the parameters from the Setup window are the same in the Multiple window. Set the desired number of measurements (laser shots). Usually, 20 measurements are satisfactory for high-signal regions of the TA spectrum. If the sample emits at the probe wavelength, be sure to check the Subtract Fluorescence Background button in the Backgrounds tab. Click on Apply then Start to collect the TA kinetics.
    NOTE: Sometimes, performing a higher number of shots (>40) shifts the baseline of the decay, positive or negative, from probe/laser scatter interference. If this is an issue, then perform a lower number of shots (~10-20) and repeat the measurement multiple times to collect multiple sets of data that can then be averaged together.
14. Once the TA measurements are complete, measure the absorption spectrum of the MOF afterward to ensure minimal degradation.

5. Narrow-angle nsTA setup

1. Sometimes with the perpendicular pump-probe setups, the signal obtained from the MOF suspension is quite weak (<10 ΔmOD) and still fluctuates from scatter due to the large sample volume being excited. To help minimize the signal fluctuations and enhance the signal, ultrafast transient absorption setups can be applied toward nsTA setups with a narrow-angle pump-probe beam orientation and smaller path lengths (Figure 2).
2. Depending on the sample chamber setup, one can focus and direct the excitation beam so that the pump and probe beams cross at angles <45°, and therefore provide more overlap. Do this with focusing optics (Figure 2, concave lens [CCL] and convex lens [CVL]) and kinematic mirrors (Figure 2, MM1-3). Turn on the laser/spectrometer system and repeat steps 4.2 and 4.3.
   NOTE: While the use of concave/convex lenses is ideal for focusing optics, an optical iris can be used in place of these components to narrow the beam. The narrowing of the beam in this manner can be compensated for by increasing the power; however, when working with wavelengths below 400 nm, degradation and bleaching of the iris is quite common. Some TA spectrometers do not have breadboards that permit mounting optics in the sample chamber. The spectrometer used here does not have breadboards, so holes were drilled and tapped in the sample chamber to set up optics (Figure 2, MM1-3). If the spectrometer is still under warranty, contact the company support team to see if they can accommodate such a setup.
3. To decrease the beam spot size hitting the 2 mm cuvette (Table of Materials), set up a Galilean telescope with a concave lens (Table of Materials, CCL1) hitting the laser first and a convex lens (Table of Materials, CVL1) hitting the laser second. Ensure that the distance between the two lenses is approximately the difference between the two focal distances of the lenses.
   NOTE: The Spectra-Physics Quanta Ray lasers used in these measurements have a spot size of 1 cm. The spot size was halved with the Galilean telescope setup. For lasers that output megawatts of power, Galilean telescopes need to be exclusively used. A Keplerian telescope (two convex lenses) forms plasma between the two lenses at even modest powers (~10 mW).
4. Open up both the laser and probe shutters. Replace the first shutter mirror (SM1) with SM2 and place a note card into the SM2 clamping mount, such that its orientation is completely facing the probe beam. Then, set up a series of mini mirrors (MM1-3), approximately as depicted in Figure 2. Direct the incoming laser beam by adjusting the turning knobs on the P3 kinematic mount approximately onto the center of MM1. To minimize laser beam expansion from mirror to mirror, place MM2 in front of MM1 to lower the angle of reflection between the two mirrors (Figure 2).
   NOTE: For laser alignments, a common practice is to adjust a mirror/prism one mirror away from the intended spot location (e.g., adjusting P2 to precisely hit MM1). However, P2 in the experiments discussed here is a stationary beam-guiding optic and should not be adjusted. If given the flexibility, alignment should be done with an optical component one mirror away from the target optic.
5. With the beam hitting approximately the center on MM1, rotate MM1 so that the reflected laser beam is hitting MM2 in the center. With the beam hitting approximately the center on MM2, rotate MM2 so that the reflected laser beam is hitting MM3 in the center. With the beam hitting approximately the center on MM3, rotate MM3 so that the reflected laser beam is hitting the alignment note card in the same spot as the probe beam.
6. Fine-tune the laser beam positions on each of the mirrors and note card with the vertical and horizontal knobs on the mirrors. Ensure that the beam has little to no clipping throughout its path.
7. Repeat steps 5.5 and 5.6 using a 2 mm cuvette with a 14/20 inner joint (SC2) and 14/20 rubber septum (Table of Materials). Insert the sample into a clamping sample mount (SM2) completely facing the probe beam path. Fine-tune the laser beam positions on each mirror and SM2 with the vertical and horizontal knobs on the mirrors.
   NOTE: For added ease in changing between perpendicular and narrow-angle TA setups, a flipping or magnetic kinematic mirror mount for MM1 can be used instead of a regular kinematic mount to avoid having to realign optics.The placement of MM2 and MM3 should not affect the incident pump or probe beams in the perpendicular setup.
8. With a low-profile stirrer (Table of Materials), stir the sample moderately and perform TA measurements. Repeat steps 4.6-4.14.
   ​NOTE: For 1-20 Hz lasers, a lower power can often be used (~1 mJ/pulse).

6. Ultrafast transient absorption measurements (ufTA)

1. Aligning pump and probe beams for maximum overlap
   1. The MOF suspension procedure in section 1 does not change. The pre-TA absorption measurements (section 2) do not change, except using SC2 instead of SC1 (Table of Materials). If needed, the purging process does not change either.
   2. To align the pump and probe beams for ufTA measurements, begin by preparing a solution of a well-known chromophore [e.g., Ru(bpy)₃²⁺] in a 2 mm path length cuvette with an OD of 0.5-1 at the excitation wavelength. There is no need to purge the sample.
      NOTE: Choose a standard sample that exhibits a TA spectrum in the same wavelength region as the MOF sample. Oftentimes, the MOF linker can be used as the standard.
   3. Turn on the ultrafast laser pump source and spectrometer (Figure 3). Open up the optical parametric amplifier software (if present) and set it to the desired excitation wavelength. Open up the ufTA spectrometer software and choose a probe window (UV-visible, visible, or near-infrared [near-IR]).
      NOTE: Make sure that the optical delay stage is aligned at short and long time delays. Depending on the system, this is done manually or through the spectrometer software. Most commercial systems have an "Align Delay Stage" option in the software that can be clicked to align it.
      NOTE: If possible, turn off the lights or minimize light interference when observing the pump and probe beams.
   4. Place the standard cuvette in the sample holder in line with the probe beam. Adjust the pump source power with an ND filter wheel (Figure 3, ufND) to see the pump beam if necessary. Place a white note card against the cuvette side facing the pump and probe beam.
   5. Adjust the pump spot on the note card with the turning knobs on the kinematic mount, such that it is at the same vertical height as the probe beam, and horizontally adjust the pump so that it is within 1 mm or 2 mm next to the probe beam. Without the note card, fine-tune the vertical and horizontal positions of the pump beam to obtain the highest TA spectral signal.
   6. Adjust the focus (Figure 3, TS) of the pump beam, so it is at its smallest spot size when hitting a standard sample cuvette. The focus is at the smallest point when the maximum signal is obtained. Once the highest spectral signal is obtained, the pump and probe beams are optimally aligned.
      NOTE: Commercial ufTA systems (Table of Materials) typically have a Live View option that allows the user to set the time zero and see the whole TA spectrum before officially measuring the sample.
2. Determining the pump beam spot size and energy density
   1. With the pump and probe beams aligned, replace the sample cell holder with a mounted pinhole wheel (2,000-25 µm holes; Table of Materials) at the focal point of the laser beam (Supplementary Figure 1, PHW). Make sure that the pinhole wheel is nearly (if not exactly) perpendicular to the path of the laser beam.
   2. Set up the pinhole wheel such that the laser beam is passing through the 2,000 µm pinhole. Set up a detector attached to a power meter (Supplementary Figure 1, PWR) closely on the other side of the pinhole wheel, such that all of the laser beam is hitting the detector.
   3. Adjust the pump source power with an ND filter wheel so the detector is measuring sufficient power. Note the average power at that pinhole size.
   4. Rotate the pinhole wheel to a smaller pinhole size and adjust the vertical and horizontal position of the laser beam to attain the maximum power output at that pinhole. Note the power for the pinhole size. Repeat this step with progressively smaller pinholes until the smallest pinhole is reached.
      NOTE: While the pinhole measurements are more of an approximate method, it is sufficient enough for measurements when comparing it to the alternative method of using a CCD camera, which can cost thousands of dollars.
   5. In data analysis software, plot the data to generate half of a pseudo-gaussian curve (it is not going to be perfect because the beam inherently is not completely gaussian). To get a symmetrical curve, take the same data and paste it in ascending order of spot sizes.
   6. Multiply the data by -1, so the minimum is now the maximum. Plot the data and fit it to a gaussian curve. Divide the maximum value of the fitted curved by e². The width of the curve at 1/e² is the approximate spot size diameter.
3. Linear power response check
   1. To ensure that no non-linear effects are present at a desired power level (e.g., multiphoton excitation processes, multiparticle decays), the signal at multiple points in the MOF TA spectrum right after the chirp response needs to be recorded at different powers. Determine five power levels to make up a curve.
   2. Replace the pinhole wheel with the sample holder and place the standard sample back in the holder. Repeat step 6.1 (the realignment process should be much easier as the pump beam was only minorly adjusted in step 6.2).
   3. Once the pump and probe beams are aligned, and the MOF sample is stirring in the sample holder, measure and record the average pump power with a power meter attached to a detector in the pump beam path.
   4. Remove the detector from the beam path, and in the Live View TA mode, record the ΔOD signal of the MOF sample at different points in the TA spectrum right after the chirp response (~2-3 ps). Repeat steps 6.3.3 and 6.3.4 at the other four power levels.
      NOTE: Sometimes the signal is quite weak at lower power levels, so if the option is available, increase the averaging time in the "live-view" mode to 5-10 s to get a better signal-to-noise ratio and lower the probe beam signal fluctuations. We commonly set the averaging time to 2-5 s throughout all power measurements and record the OD at a wavelength with each subsequent averaging period a few times to get a standard deviation at each power.
   5. Plot the recorded data points as ΔOD versus incident power in data analysis software. If there is a linear power response, the resulting plot forms a straight line, with the y-intercept at zero. If there is a non-linear power response, as expected, significant deviations from a linear curve are typically observed.
4. Determining the energy density hitting the suspension sample
   1. With the pump beam spot size and incident power hitting the MOF suspension known, the approximate energy density can be determined.
      NOTE: For example, an approximate spot diameter of 250 µm provides a radius of ~125 µm. After converting the radius to cm, the surface area of the spot can be calculated: A = πr² = π(0.0125cm)² ≈ 0.0005 cm². Dividing the incident power (e.g., 30 µW) by the laser repetition rate (500 Hz) gives an average energy per pulse of 0.06 µJ. Finally, by dividing the average energy per pulse with the spot surface area, an average energy density per pulse of 120 µJ·cm^-2 is obtained. The ideal energy density is one that provides an adequate TA signal while falling in a linear range of pump power; however, if a lower power can be used without sacrificing too much signal, it should be used. A ΔmOD of ~1 at <10 ps is a good compromise between signal and pump power.
5. Performing ultrafast TA measurements
   1. With the MOF sample in the holder, pump and probe beams overlapped, and an ideal excitation power chosen for the sample, perform ufTA measurements.
   2. Check the Live View window and ensure the time zero is correctly set to the start of the detector chirp.
      NOTE: When switching between the standard sample and the MOF sample, the time zero may be slightly shifted, hence the need to check again.
   3. Exit out of the Live View window to the main spectrometer software. Ensure that the MOF suspension provides an optimal TA spectrum throughout the scanned time window by setting parameters for a quick scan and clicking the Start button. Typical quick-scan parameters are a time window of -5 ps to 8,000 ps, one scan, 100 data points, an exponential point map (i.e., the 100 data points recorded in increments that fit an exponential curve), and an integration time of 0.1 s.
   4. Once the quick-scan ufTA spectrum is finished and looks good overall, change the scan parameters for a higher-quality measurement and click the Start button. Typical parameters are a time window of -5 ps to 8,000 ps, three scans, 200-300 data points, an exponential point map, and an integration time of 2-3 s.
      ​NOTE: It is generally advised that the measurement time should not exceed 1 h to avoid prolonged degradation, especially at higher pump powers.
   5. Once the high-quality ufTA spectrum is finished, take the sample out of the sample holder and measure the absorption spectrum of the sample to ensure little degradation. Further confirm minimal degradation by passing the suspension through a 20 nm syringe filter (Table of Materials) and measure the absorption spectrum again.

7. Preparation of MOFs for emission measurements

1. Depending on the excitation wavelength, PNH₂ emits fluorescence and is consequently omitted from this procedure to obtain the true emission spectra and kinetics of the MOF suspension. Additionally, the syringe filtration process in steps 1.7 and 1.8 is omitted.
   NOTE: These omissions do not noticeably impact emission measurements.
2. Weigh out 1 mg of MOF and transfer it to a clean vial. Transfer 3-4 mL of DMF to the vial containing MOF. Repeat step 1.3.
3. Measure the absorbance spectrum of the MOF suspension and dilute the suspension until an OD of 0.1-0.2 is attained at the excitation wavelength (section 2).
4. Perform the aforementioned purging procedure (section 3). The MOF suspension is now ready for fluorescence measurements.

8. MOF emission measurements

1. Turn on the fluorimeter and arc lamp (Table of Materials, Supplementary Figure 2). Open the fluorimeter software and select the emission mode. Place the purged MOF suspension into the sample holder and moderately stir.
2. With the excitation wavelength established in step 7.3, set the excitation and emission monochromator slits to 5 nm as a starting point and perform a cursory emission scan with an integration time of 0.1 s.
3. Once the emission bandwidths have been optimized to deliver a good signal (>10,000 counts), measure the MOF emission spectrum using a 1 s integration time (or longer). Then, measure the excitation spectrum of the MOF at a selected emission wavelength. Ensure that the excitation spectrum looks nearly identical to the MOF absorption spectrum.
4. Close the arc lamp slit and switch the instrument mode to TCSPC (time-correlated single photon counting) on the software.
5. Select one of the LEDs used for TCSPC with the desired excitation wavelength and attach it to a sample chamber window perpendicular to the detector window. Attach the necessary wires to the LED to integrate it into the fluorimeter.
   1. Set the instrument to the desired emission wavelength, the bandwidth to 5 nm (adjust if necessary), and the time window to 150 ns as a starting point (it can be shortened depending on the sample lifetime). Apply these settings and start the TCSPC measurements from the software window.
      NOTE: A general stopping point for most TCSPC measurements is when the maximum counts have reached a value of 10,000. Additionally, the optimal detector counting rate is 1%-5% of the LED repetition rate in order to follow Poisson statistics. Consult the TCSPC LED manufacturer to obtain the device specifications if not already provided.

Results

The electronic absorption spectra of PCN-222(fb) with and without PNH₂ and filtering are shown in Figure 4. The MOF without PNH₂ was just tip-sonicated and diluted. When comparing the two spectra, the biggest difference is the minimization of baseline scatter, which shows up as a broad upward absorption with decreasing wavelengths and also broadens the electronic transitions quite noticeably. For further comparison, the PCN-222(fb) ligand in solution, tetracarboxyphenyl...

Discussion

While the above results and protocol delineate general guidelines for minimizing scatter from MOFs in spectroscopic characterization, there is a wide variability in MOF particle size and structure that impacts spectroscopic results, and therefore blurs the methods of interpretation. To help clarify interpretation and ease the strain that comes with analyzing MOF spectroscopic data, finding a procedure to make the MOFs as small as possible is key. This is a limiting factor for most spectroscopy-related analyses of MOFs. B...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
1 cm cuvette sample mount (SM1)	Edinburgh Instruments	n/a	Contact company
1 mL disposable syringes	EXELINT	26044
10 mL disposable syringes	EXELINT	26252
1-dram vials	FisherSci	CG490001
20 nm syringe filters	VWR	28138-005	The filters are made by Whatman/Cytiva, and their catalog number is 6809-1002
200 nm syringe filters	Cytiva, Whatman	6784-1302
Absorption spectrophotometer	Agilent	Cary 5000 Spectrophotometer	Contact company
Acetronitrile (ACN)	FisherSci	AA36423
Ar gas tank	Linde/PraxAir	P-4563
bis amino-terminated polyethylene glycol (PNH2)	Sigma-Aldrich	452572	MOF suspending agent
Clamping sample mount for nsTA (SM2)	Ultrafast Systems	n/a	Contact company
Concave lens for telescope(CCL1)	Thorlabs	LD1613-A-ML
Convex lens for telescope (CVL1)	Thorlabs	LA1708-A-ML
Custom 1 cm optical cell with 24/40 outer joint	QuarkGlass	QSE-1Q10-2440 (Spectrosil Cat #1-Q-10	We requested the 1 cm cell to have a joint
Custom 2mm optical cell with 14/20 outer joint	QuarkGlass	QSE-1Q2-1420 (Spectrosil Cat # 1-Q-2)	We requested the 2 mm cell to have a joint
Dimethylformamide (DMF)	FisherSci	D119
Dye laser (Nd:YAG pumped) for 415 nm output	Sirah	CobraStretch
Dye laser dye, Exalite 417	Luxottica	4170
Femtosecond laser	Coherent	Astrella
Fluorimeter	Photon Technology Inc. (Horiba)	QuantaMaster QM-200-4E
Fluorimeter arc lamp, 75 W	Newport	6251NS
Fluorimeter PMT	Hamamatsu	1527
Fluorimeter Software	PTI/Horiba	FelixGX
Fluorimeter TCSPC Module	Becker & Hickl GmbH	PMH-100
lens mounts for telescope	Thorlabs	LMR1
Long purging needles	STERiJECT	PRE-22100
Magnetic stirrer	Ultrafast Systems	n/a	Contact company
mirror 1 (MM1) 350-700 nm	Newport	10Q20BB.1
MM1 mount	Thorlabs	KM100
MM1 post	Thorlabs	TR2
MM1 post holder	Thorlabs	PH1.5
MM2 mount	Thorlabs	MFM05
MM2,3 mirrors	thorlabs	BB03-E02
MM2,3 post	Thorlabs	MS3R
MM2,3 post bases	Thorlabs	MBA1
MM2,3 post holders	Thorlabs	MPH50
MM3 mount	Thorlabs	MK05
mounting posts for telescope optics	Thorlabs	TR4
Nanosecond TA Nd:YAG lasers	Spectra-Physics	QuantaRay INDI Nd:YAG
Nanosecond TA spectrometer	Edinburgh Instruments	LP980
nsTA ICCD camera	Oxford Instruments	Andor iStar ICCD camera	Contact company
nsTA PMT	Hamamatsu	R928
Optical parametric amplifier	Ultrafast Systems	Apollo
Parafilm	FisherSci	S37440
Pinhole wheel	Thorlabs	PHW16
Pinhole wheel post base	Thorlabs	CF125C
Pinhole wheel post holder	Thorlabs	PH1.5
Pinhole wheel post/mount assembly	Thorlabs	NDC-PM
post bases for telescope optics	Thorlabs	CF125C
post holders for telescope optics	Thorlabs	PH4
Power detector for ns TA	Thorlabs	S310C
Prism assembly (P2,3)	Edinburgh Instruments	n/a	Contact company
Prism mount (P1)	OWIS	K50-FGS
Prism post (P1)	Thorlabs	TR4
Prism post base (P1)	Thorlabs	CF125C
Prism post holder (P1)	Thorlabs	PH4
Quartz prisms (P1-P3)	Newport	10SR20
Rubber outer joint septa (14/20)	VWR	89097-540
Rubber outer joint septa (24/40)	ChemGlass	CG-3022-24
Sonication tip	Branson	product discontinued	Closest alternative is 1/8" diam. tip from iUltrasonic
Square ND filters	Thorlabs	NEK01S
Stir bars	StarnaCells/FisherSci	NC9126395
Thorlabs power detector for ufTA	Thorlabs	S401C
Thorlabs power meter	Thorlabs	PM100D
Tip sonicator	Branson	Digital Sonifer 450, product discontinued	Closest alternative is SFX550 from iUltrasonic
Tygon tubing	Grainger	8Y589
ufTA ND filter wheel	Thorlabs	NDC-25C-2-A
ufTA ND filter wheel mount	Thorlabs	NDC-PM
ufTA ND filter wheel post	Thorlabs	PH2
ufTA ND filter wheel post base	Thorlabs	CF125C
ufTA pump alignment mirror	Thorlabs	PF10-03-F01
Ultrafast TA telescope assembly	Ultrafast Systems	n/a	Contact company
Ultrafast transient absorption spectrometer	Ultrafast Systems	HeliosFire
Xe arc probe lamp	OSRAM	4050300508788