Tuning the Performance of Metalloporphyrin-Based Pressure-Sensitive Paints through the Nature of the Central Metal Ion

Abstract

Since the 1980s, pressure-sensitive paint (PSP) has been used as an optical pressure sensor for measuring surface pressure on aircraft models in wind tunnels. Typically, PSPs have utilized platinum(II)-5,10,15,20-tetrakis(2,3,4,5,6-pentafluorophenyl)-porphyrin due to its high pressure sensitivity, phosphorescence lifetime of ∼50 μs, reasonable quantum yield of emission, and resistance to photo-oxidation. This work investigates the photophysics and electronic structure of metal complexes of 5,10,15,20-tetrakis(2,3,4,5,6-pentafluorophenyl)-porphyrin, namely, Zn(II), Pd(II), and Ir(III), as potentially improved luminophores for polymer-based PSPs. The metal ion was found to preferentially stabilize the a_2u MO of the porphyrin with increasing electronegativity, thus blue-shifting absorption/emission maxima and increasing Q-band intensity. The lifetime and quantum yield of emission increased and decreased, respectively, in the order of Pt(II) to Ir(III) to Pd(II), primarily due to the heavy atom effect. The increase in phosphorescence lifetimes resulted in the pressure sensitivity of the PSPs increasing in the order of Pt(II) to Ir(III) to Pd(II). However, the temperature sensitivity at pressures >70 kPa also increased with increasing phosphorescence lifetime. Overall, this work identified that the central metal ion of porphyrin luminophores can be used to tailor the resulting lifetime of the luminophore and therefore heavily influences the pressure and temperature sensitivity of polymer PSP formulations. This new insight into luminophore design can be used to optimize PSPs for a desired application.

Introduction

Wind tunnel tests of aerospace vehicles and aircraft are increasingly making use of pressure-sensitive paint (PSP). PSP is an optical oxygen sensing technique that involves the formulation of an oxygen-sensitive phosphorescent compound, known as a luminophore, and an oxygen-permeable binder into a paint, which is then applied to the surface of a desired test model.¹ When this coated model is placed into a wind tunnel, the static air pressure distribution causes different levels of oxygen diffusion into the binder at different locations. If light of an appropriate wavelength illuminates the model, the luminophore is excited into a singlet excited state and, through intersystem crossing, can be converted to a long-lived triplet excited state (T₁). From the T₁ excited state, the luminophore can relax back to the ground state via release of a lower energy photon, through a process called phosphorescence. The phosphorescent intensity is dependent on the local oxygen concentration and therefore, can be imaged and related to the surface pressure through in situ or a priori calibration.¹ The PSP method offers advantages over traditional aerodynamic measurements, such as pressure taps, because of its global resolution and nonintrusive application. Work over the past 30 years has demonstrated the efficacy of PSP in accurately studying various aerodynamic applications such as steady and unsteady flows,²⁻⁹ rotating flows,¹⁰⁻¹² cryogenic experiments,¹³⁻¹⁵ and even blast waves.^16,17

One of the main drawbacks concerning PSP measurements is the inherent temperature sensitivity, resulting in inaccurate pressure determination. This results from nonradiative decay of the luminophore T₁ state and temperature-dependent oxygen diffusion through the paint binder.¹ Attempts to resolve the temperature sensitivity issue have involved the development of dual-luminophore PSPs that contain a pressure-sensitive luminophore and a secondary temperature-sensitive but pressure-insensitive luminophore. The secondary signal can then be used to account for errors in temperature change associated with the primary luminophore signal.^5,18,19 However, some dual-luminophore formulations suffer from luminophore cross-talk which can hinder PSP performance.^20,21

PSP development in recent years has largely focused on binder development, such as new polymers,²²⁻²⁶ poly ceramic PSP (PC-PSP),^27,28 and anodized aluminum PSP (AA-PSP),²⁹⁻³¹ in an attempt to improve response times for unsteady flow measurements. However, a distinct lack of development of the luminophore species has been pursued. Luminophore molecular design can be used to tailor the lifetime/quantum yield of emission and the molar absorptivity of the resulting PSP, thus affecting properties like pressure sensitivity S_p, temperature sensitivity, S_T, and the brightness of emission. Additionally, the solubility of a luminophore, and its interaction with the binder matrix, can greatly impact an optical molecular-based sensor’s performance.³² Traditionally, PSP formulations have used Pt(II) porphyrins, Ru(II) polypyridyls, and pyrene derivatives as the luminophore.¹ Ru(II) polypyridyls have high quantum yields and adsorb well onto porous binders.³³ However, Pt(II) porphyrin-based PSPs typically have a higher S_p and a relatively lower S_T but do not adsorb well onto porous binders. For traditional polymer-based PSPs, the luminophore Pt(II)-5,10,15,20-tetrakis(2,3,4,5,6-pentafluorphenyl)-porphyrin (PtTFPP) is most ubiquitously used due to its high degree of photostability and solubility in many common PSP polymers.³⁴ Indeed, the only commercially available PSPs utilize PtTFPP as the luminophore.³⁵ We recently explored the effect of tetraphenyl porphyrin degree, type, and position of phenyl halogenation on PSP performance and found these factors can have a large effect on the subsequent polymer-based PSP performance.³⁶ This was especially true with S_T, finding that PtTFPP polystyrene PSPs possessed the highest S_T. Through luminophore design, we demonstrated that S_T can be reduced by 0.5%/K. However, the S_p in comparison was largely unaffected by porphyrin phenyl halogenation due to their similar lifetimes in the polymer binder matrix. Considering these results, this study investigates the effect of the central metal ion on the photophysics of freebase and metal complexes of TFPP and the resulting effect on polymer-based PSP performance. Typical phosphorescent metalloporphyrins constitute Pt(II), Pd(II), and Ir(III). Therefore, we introduce Ir(III)(Cl)(CO)TFPP (Ir1) for the first time in PSP formulations along with PtTFPP (Pt1) and PdTFPP (Pd1) to explore the effect of the central metal ion on the PSP performance (Figure 1). Additionally, Zn(II)TFPP (Zn1) and the freebase porphyrin, TFPP (Fb1), were included to study fluorescent porphyrins in polymer-based PSPs.

Figure 1.

Chemical structures of the freebase porphyrin Fb1 and the metalloporphyrins Zn1, Ir1, Pd1, and Pt1.

Experimental Section

Detailed synthetic procedures and characterization data can be found in the accompanying Supporting Information.

Photophysical Studies

UV–vis electronic absorption spectra were recorded on a Mettler Toledo UV5Bio spectrophotometer. Steady-state emission and excitation spectra and lifetime data were recorded on an Edinburgh Instruments FP920 Phosphorescence Lifetime Spectrometer equipped with a 450 W steady state xenon lamp, a 5 W microsecond pulsed xenon flash lamp (with single 300 mm focal length excitation and emission monochromators in Czerny Turner configuration), interchangeable EPL pulsed diode lasers, and a red-sensitive photomultiplier in Peltier (air cooled) 53 housing (Hamamatsu R928P). Plotting, fitting, and analysis of data were carried out using Origin 2019b. All data were fitted with exponential decay models and the goodness of fit evaluated by residual, χ², and R² analysis. These measurements were recorded in chloroform which was dried over 4 Å molecular sieves and degassed using three freeze–pump–thaw cycles and standard Schlenk techniques. All samples were prepared in an Innovative Technologies System Two, under argon, where the concentration of oxygen and water was always kept below 0.1 ppm. Emission spectra and lifetime measurements were recorded in J. Young’s taps sealed quartz cuvettes. Quantum yields were calculated using the relative method, with tetraphenyl porphyrin in toluene (Φ = 0.07) as a standard. Polystyrene-doped samples were prepared by drop casting approximately 100 μL of the PSP polystyrene solution in chloroform onto a glass microscope slide and air-dried. These samples were subsequently taken into an argon-filled glovebox to remove all residual volatiles and sealed with another glass slide using vacuum grease around the edges to prevent diffusion of oxygen into the sample. The reported data are an average of three independent measurements. The weight-averaged emission lifetimes in argon-saturated polystyrene, ⟨τ⟩ were calculated using eq 1.

1

where B is the initial intensity of a given component and τ is the lifetime of a given component.

PSP Recipes

The polystyrene-based PSPs were formulated by dissolving 1 g of polystyrene (Sigma-Aldrich M_w ∼ 380,000) in 25 mL of chloroform and using a luminophore loading of 0.68% wt/wt luminophore/polystyrene.

PSP Performance Studies

PSP formulations were sprayed onto Ambersil matt white RAL 9010 base-coated aluminum coupons using a spray gun in 12 light coats. The samples were left to air-dry for 30 minutes after spraying.

The performance of the PSP formulations was investigated in the standard approach of a priori calibration using the University of Manchester PSP calibration chamber. The a priori calibration was performed using the published procedure.³⁶

The pressure sensitivity at a certain temperature, S_P(T) was calculated from the slope of the modified Stern–Volmer calibration plots using 2 with I_ref and P_ref as the luminescence intensity and pressure, respectively, at 100 kPa and 293 K.

2

The temperature sensitivity at a given pressure, S_T(P), is calculated as the percentage change in I_ref/I with respect to the temperature by using 2 with I_ref as the luminescent intensity at 100 kPa and 293 K.

3

Photodegradation studies were conducted at room temperature and pressure and involved recording images of the PSP luminescence every 3 min for 45 min of constant illumination. The amount of photodegradation was calculated as the % luminescent intensity (at 45 min) of the original luminescent intensity (at 0 min).

Results and Discussion

Synthesis

The freebase porphyrin Fb1 was synthesized via the standard Lindsey conditions in a 19% yield.³⁷ The Zn(II) complex (Zn1) was synthesized by heating Fb1 with Zn(OAc)₂ to reflux in a chloroform–methanol mixture, giving an isolated yield of 86%. The Ir(III) complex (Ir1) was synthesized by heating Fb1 with Ir(COD)₂Cl₂ to reflux in 1,2,4-trichlorobenzne, affording an isolated yield of 46%.³⁸ The Pd(II) and Pt(II) (Pd1 and Pt1) complexes were synthesized by heating PdCl₂/PtCl₂ to reflux in benzonitrile under argon, before adding Fb1, resulting in isolated yields of 92% and 87%, respectively (see the accompanying Supporting Information for full synthetic procedure and characterization data).³⁶

Single-Crystal X-ray Crystallography

Suitable crystals for single-crystal XRD were grown through slow evaporation of a 1:1 DCM/hexane mixture for Zn1, Pd1 and 1:1 DCM/chloroform for Ir1. The single-crystal XRD structure for Pt1 has been previously published (CCD deposition number: 2338068).³⁶ Normal coordinate structural decomposition (NSD) was used to evaluate the in-plane (IP) and out-of-plane (OOP) distortions adopted by the different porphyrins in their single-crystal geometries. NSD analysis was performed using the web-based program developed by Kingsbury and Senge.³⁹ The structures of the metalloporphyrins (Figure 2) are all relatively flat, with Pt1, Pd1, and Ir1 all having a small Δoop = 0.03 Å and the metal ion sitting in the plane of the 4 nitrogen atoms (Table 1). Pt1 and Pd1 adopt a slight wave (E_gx/E_gy), while Ir1 adopts a slight dome (A_2u) conformational distortion. Zn1 has a much larger Δoop = 0.15 Å than the other metalloporphyrins, adopting a wave (E_gx/E_gy) conformational distortion. The porphyrins also possess IP distortions, which are expansions of the porphyrin core around the central metal ion. Zn1 has the highest Δip = 0.19 Å and a Zn–N bond distance of 2.044(14) Å, which is consistent with other Zn(II) porphyrins, Zn–N(range) = 2.298(7) – 1.799(8) Å.⁴⁰ Zn(II) has low-energy 3d orbitals, which are unable to π-backbond with the high-energy porphyrin π* orbitals; therefore, a weak and long Zn–N bond is formed. Looking at Pd1, the Δip = 0.08 Å and the Pd–N bond distance = 2.017(16) Å, which is consistent with other Pd(II) porphyrins, Pd–N(range) = 2.131(4) – 1.926(12) Å.⁴⁰ The Pd(II) ion forms a stronger M–N interaction because of the higher in energy 4d orbitals, and thus greater π-backbonding can occur. Pt1 has a Δip = 0.07 Å and a Pt–N bond distance of 2.009(2) Å, which is consistent with other Pt(II) porphyrins, Pt–N(range) = 2.069(4) – 1.960(10) Å.⁴⁰ For the Pt(II) porphyrin, the M–N interaction is even stronger because the Pt(II) ion has even higher energy 5d orbitals, which facilitates even stronger π-backbonding. The Ir–N bond in Ir1 is markedly longer at 2.050(3) Å.

Figure 2.

Solid-state structures of (top) Zn1 and Ir1 (bottom) Pd1. Thermal ellipsoid probability set at 50% with hydrogens omitted for clarity. Colors are assigned: C = gray, N = blue, O = red, F = light green, Cl = dark green, Zn = pink, Ir = orange, Pd = purple.

Table 1. Key NSD and Structural Parameters of the Solid-State Structures of the Metalloporphyrins.

porphyrin	out-of-plane deformation (Å)						distortion parameters (Å)		structural parameters (Å)
	B2u	B1u	A2u	Eg(x)	Eg(y)	A1(u)	Δoop	Δip	M–N bond distance
Zn1	0	0	0	0.14	0.06	0	0.15	0.19	2.044(14)
Ir1	0	0	0.02	0	0	0.02	0.03	0.18	2.050(3)
Pd1	0	0	0	0.03	0.02	0	0.03	0.08	2.017(16)
Pt1a	0	0	0	0.01	0.02	0	0.03	0.07	2.009(2)

^aData previously published.³⁶

UV–Vis Electronic Absorption Spectroscopy

The UV–vis electronic absorption spectra of all metalloporphyrins in chloroform were found to be characteristic of typical metalloporphyrins, with an intense Soret band (B band) around 400 nm and two much weaker in intensity Q bands centered on 550 nm (Figure 3 and Table 2).

Figure 3.

UV–vis absorption spectra of Fb1, Zn1, Ir1, Pd1, and Pt1 in chloroform at a concentration of 2 μM. All values are normalized to the Soret band maximum. Inset shows the zoomed-in Q-band region of 500 to 675 nm. A simplified energy-level diagram showing the four Frontier orbitals of the porphyrins, and the principal electronic transitions that form the Soret and Q bands is also provided.

Table 2. Soret and Q Band Maxima with Molar Absorption Coefficients, ε in M^–1 cm^–1 in Parentheses, of the Porphyrins in Chloroform^a.

porphyrin	Soret band maximum (nm) [ε M–1 cm–1]	Q-band maxima (nm) [ε M–1 cm–1]				integrated intensity of Q(0,0)/Q(1,0)
		Q1	Q2	Q3	Q4
Fb1	412 [314,900]	507 [22,300]	537 [2820]	583 [7100]	637 [1200]	0.12
Zn1	414 [468,600]	524 [22,200]	578 [3000]			0.25
Ir1	414 [176,800]	527 [19,100]	558 [6000 ]			0.34
Pd1	407 [271,400]	519 [22,200]	552 [18,800]			0.53
Pt1b	392 [296,500]	507 [18,800]	540 [28,400]			0.79

^aThe ratio of the integrated intensity of Q_abs(0,0) to the integrated intensity of Q_abs(1,0) is also presented (for Fb1, this is calculated as [Q_y(0,0) + Q_x(0,0)/Q_y(1,0) + Q_x(1,0)]).

^bData previously published.³⁶

The spectral features can be explained using Gouterman’s four orbital model, where the four Frontier orbitals: highest occupied molecular orbital (HOMO)–1 (a_1u), HOMO (a_2u), and two degenerate lowest unoccupied molecular orbitals (LUMOs) (e_g) are relatively isolated in energy.⁴¹ The Soret band is formed predominantly of a π–π* transition from HOMO–1 to the degenerate LUMOs. Whereas the Q bands are formed of Q_abs(1,0) and Q_abs(0,0), which are assigned as a vibronic satellite and the electronic origin π–π* transitions from the HOMO to the degenerate LUMOs. Freebase porphyrins (Fb) are known to have a further split in each Q-band due to the symmetry break of the central proton axis. Consequently, for Fb1, there are two sets of Q bands, Q_y and Q_x, along with their accompanying vibrational satellites, resulting in four Q bands in total. The position of the Soret band blue-shifts in the order of Ir1 = 414 nm and Zn1 = 414 nm > Fb1 = 412 nm > Pd1 = 407 nm > Pt1 = 392 nm. The reason for this blueshift has recently been revisited with Ghosh et al. finding that the metal ions stabilize the a_2u HOMO–1 increasing the HOMO–1 to LUMOs energy gap and thus the Soret band energy.⁴² The intensity of the Q_abs(0,0) band increases across the family of porphyrins. In contrast, the intensity of the vibronic satellite, Q_abs(1,0), remains roughly consistent because it derives its absorption strength from the B transitions (that make up the Soret band) via Herzberg–Teller coupling, which exceeds the intensity from the normal Franck–Condon progressions.⁴³ The increasing intensity of Q_abs(0,0) can best be seen using the ratio of the integrated intensity of Q_abs(0,0) to Q_abs(1,0) band (because the intensity of the Q(1,0) band remains roughly constant), which increases in the order of Fb1 = 0.12 < Zn1 = 0.25 < Ir1 = 0.34 < Pd1 = 0.53 < Pt1 = 0.79. For the heavily fluorinated porphyrin (TFPP) used in this study, the a_2u MO is already greatly stabilized and so the freebase porphyrin, Fb1, has its a_2u MO and a_1u MO flipped. Therefore, the a_2u becomes the HOMO–1 and the a_1u the HOMO.⁴⁴ As metal ions are incorporated into this porphyrin, the a_2u HOMO–1 will be increasingly stabilized, with the strength of this stabilization dependent on the electronegativity of the metal. Therefore, any metal ion will stabilize the a_2u HOMO–1 away from the a_1u HOMO, decreasing orbital mixing and increasing Q_abs(0,0) intensity. The exception to this trend is Ir1, which has a reduced intensity Q_abs(0,0) band when compared to Pd1, despite Pd(II) and Ir(III) having similar electronegativity values (2.70 and 2.79, respectively).⁴⁵ The axial Cl and CO act as electron-withdrawing substituents, which can lower the energy of the a_1u HOMO, thus bringing it closer to the a_2u HOMO–1, therefore, increasing the level of orbital mixing, which in turn reduces Q_abs(0,0) intensity.

Density Functional Theory and Time-Dependent-Density Functional Theory Calculations

To further explore the electronic structure of Fb1, Zn1, Ir1, Pd1, and Pt1, DFT and time-dependent DFT (TD-DFT) calculations were carried out. Calculations employed either the B3LYP⁴⁶ (DFT) or CAM-B3LYP⁴⁷ (TD-DFT) functional with the lanl2dz⁴⁸ basis set for all atoms. Solvation in chloroform was treated with a conductor-like polarizable continuum model and the D4 charge-dependent atom-pairwise dispersion correction⁴⁹ was included throughout. This methodology has been found to be optimal in previous porphyrin DFT and TD-DFT calculations.⁵⁰ All structures are optimized and confirmed as minima on the potential energy surface, through the absence of imaginary vibrational modes. All calculations were performed by using the Orca 5.0.4 software package.

DFT Calculations

All calculated structures are near-flat. The relative energies of HOMO–1, HOMO, LUMO, and LUMO+1, with the accompanying relevant energy gaps, are found in Figure 4 and Table S1. The porphyrins possess an energetically isolated degenerate LUMO and LUMO+1 (except Fb1, which has the LUMOs split in energy by 0.01 eV, due to the central proton axis breaking symmetry) and a close in energy HOMO and HOMO–1. The HOMO and HOMO–1 are a_2u and a_1u π MO and are identical across all of the systems studied.

Figure 4.

Energy-level diagram showing the energy, in eV, of the HOMOs (bottom): a_1u (blue) and a_2u (red) and the LUMOs (top) of Fb1, Zn1, Pt1, Pd1, and Ir1. Orbitals are drawn with an isosurface value of 0.03 e/A³.

Comparing Fb1 to Zn1, the incorporation of the Zn(II) ion raises the HOMO and HOMO–1 in energy, and they also become closer in energy, with HOMO to HOMO–1 gaps of 0.11 and 0.03 eV, respectively. For Pt1, Pd1, and Ir1, the DFT predicts that the a_2u MO is increasingly stabilized by the increasingly electronegative metal ions, thus causing the a_2u MO to become the HOMO–1 as it is stabilized below the a_1u MO in Pd1 and Pt1. Therefore, the HOMO to HOMO–1 energy gap increases in the order of Ir1 = 0.01 eV < Pd1 = 0.08 eV < Pt1 = 0.13 eV. The a_2u MO of Pt1 and Ir1 are predicted to be at similar energies (−6.40 and −6.41 eV, respectively), owing to the relatively similar electronegativities of the metal ions (2.98 and 2.79, respectively). However, for Ir1 the a_1u MO is greatly stabilized and the a_2u MO is slightly stabilized. This makes the two MOs almost degenerate in energy, contributing to the much smaller HOMO to HOMO–1 energy gap for Ir1. The average of the HOMO–1 to LUMO and HOMO to LUMO energy gaps affords the center of gravity of the resulting absorption spectrum, which then determines Soret band energies. The center of gravity of the absorption spectrum is predicted to increase steadily in the order of Fb1 = 2.89 eV < Zn1 = 2.93 eV < Ir1 = 3.04 eV < Pd1 = 3.08 eV < Pt1 = 3.16 eV because the a_2u MO is being stabilized away from the a_1u MO. The exception for this is Ir1, which has a similar a_2u MO energy to Pt1, but its degenerate LUMOs are predicted to be stabilized by about 0.2 eV with respect to the other metalloporphyrins, leading to smaller HOMOs to LUMOs energy gaps.

TD-DFT Calculations

TD-DFT calculations predict the UV–vis electronic spectra generally well, with each porphyrin having four major excited states, except for Pd1 (Table S3), in line with the four-orbital model. The absorption spectra of the metalloporphyrins are predicted to have Soret and Q bands, each made of two symmetry matched excited states, B_y/B_x and Q_y/Q_x, respectively (with B_y and B_x forming the Soret band and Q_y and Q_x forming the Q_abs(0,0) feature). The freebase porphyrin, Fb1, is predicted to have two symmetry split excited states for each absorption feature due to the central proton axis energetically splitting the LUMO and LUMO+1. Surprisingly, the Soret band feature of Pd1 is predicted to be formed of three degenerate excited states. Across the series, Soret bands are predicted to blue-shift in wavelength, which matches the experimental data. This blueshift is due to the increasing energy gap between the LUMOs and the center of gravity of the absorption spectrum, with increasing metal ion electronegativity. However, the exception is Ir1 with its lower in energy LUMOs, which decrease the HOMOs to LUMOs energy gap. Therefore, Ir1 possesses a red-shifted Soret band, with respect to Pd1 and Pt1. The oscillator strength of the Q-band feature is predicted to increase steadily from Fb1 ≈ 0 < Zn1 = 0.001 < Ir1 = 0.003 < Pd1 = 0.010 < Pt9 = 0.021. This trend aligns with the experimental data, where the intensity of the Q_abs(0,0) band increased across the series. The four orbital model predicts that each excited state is made up of a pair of one-electron transitions from the HOMO–1 and HOMO to the LUMOs.⁴¹ Different combinations of these one-electron transitions can happen constructively, leading to the intense Soret band feature, or destructively, leading to the much weaker Q-band features.⁵¹ The more mixing of the electron transitions, the more destructive combinations are possible and thus the weaker the Q bands. This orbital mixing is dependent on the energy gap between the HOMO and HOMO–1 orbitals in the ground state, which is most evident when comparing the sum of the predicted contributions of the transitions from the HOMO–1 to the LUMOs and the HOMO to the LUMOs, for a given excited state. Focusing on the Q_x excited state, for example, (Figure 5 and Table S3), the sum of the contributions of the HOMO to LUMOs transitions become increasingly more dominant over the sum of the contributions of the HOMO–1 to LUMOs transitions in the Q_x excited-state makeup across the series.

Figure 5.

Sum of the predicted % contribution of the transitions from the HOMO–1 and HOMO to the LUMOs for the Q_x excited state from the TD-DFT calculations.

The increasing dominance of the HOMO to LUMOs transitions (and therefore, decreasing dominance of HOMO–1 to LUMOs transitions) in the Q_x excited-state makeup, follows the progression Fb1 = 53% ≈ Zn1 = 52% ≈ Ir1 = 53% < Pd1 = 59% < Pt1 = 62%. The HOMO to LUMOs transitions become increasingly dominant in the excited-state composition from Zn1 to Pd1 to Pt1 due to the increasing stabilization of the a_2u MO, resulting in a greater HOMO to HOMO–1 energy gap. However, for Ir1 the a_1u MO is also stabilized, which decreases the HOMO to HOMO–1 energy gap. As less orbital mixing occurs in the excited state, the Q bands increase in intensity because fewer destructive combinations are possible. It is useful to look at the resulting MO energy levels in the excited structures to better explain the excited-state composition (Table S2). Upon excitation, the now singly occupied molecular orbital (SOMO)–LUMO/LUMO+1 destabilize in energy and the now SOMO–HOMO/HOMO–1 stabilize in energy. The a_2u MO stabilizes in energy to a greater extent than the a_1u MO, causing the a_2u MO to become the HOMO–1 in all metalloporphyrins. Whereas in Fb1, the a_1u MO remains as the HOMO–1. Consequently, the HOMO to HOMO–1 energy gaps in the excited structures are Fb1 = 0.01 eV < Zn1 = 0.06 eV < Ir1 = 0.08 eV < Pd1 = 0.16 eV < Pt9 = 0.22 eV, which matches the experimentally observed increasing intensity of the Q_abs(0,0) feature across this series.

Emission Spectroscopy

The emission spectra of the porphyrins in chloroform possess 2 distinct bands, the electronic origin of the emission, denoted as Q_em(0,0) and a lower in energy vibronic satellite, denoted as Q_em(0,1) (Figure 6 and Table 3). The Q_em(0,0) maxima follow the same trend as the Soret maxima with Ir1 = 680 nm > Pd1 = 672 nm > Pt1 = 651 nm. However, the Q_em(0,0) maximum of Zn1 is blue-shifted, with respect to the other metalloporphyrins to 599 nm, which is common for Zn(II) porphyrins.⁵² Similar to the absorption spectra, the Q_em(0,0) feature increases in intensity with increasing metal ion electronegativity, for example, the integrated intensity of Q_em(0,0)/Q_em(0,1), increases in the order of Fb1 = 0.26 > Zn1 = 0.16 < Ir1 = 0.98 < Pd1 = 1.30 < Pt1 = 1.76. The increasing intensity of the spectral features suggests that the HOMO to HOMO–1 energy gap is increasing across this series.

Figure 6.

(Left) Emission spectra of Fb1, Zn1, Ir1, Pd1, and Pt1 in deoxygenated chloroform at a concentration of 0.5 μM. All values are normalized to the Q_em(0,1) maximum of the respective porphyrin to highlight increasing Q_em(0,0) intensity. Excitation was in the Soret band maximum of the respective porphyrin. (Right) Lifetimes of emission for Ir1, Pd1, and Pt1 in deoxygenated chloroform at a concentration of 0.5 μM, fitted to a monoexponential decay. Inset shows a zoomed in section to highlight the decays of Ir1 and Pt1. The lifetimes and associated fits for Fb1 and Zn1 can be found in the accompanying Supporting Information.

Table 3. Emission Peak Maxima, Lifetimes of Emission, and Quantum Yields of Emission in Deoxygenated Chloroform^a.

porphyrin	Qem(0,0) maximum (nm)	Qem(0,1) maximum (nm)	integrated intensity of Q(0,0)/Q(0,1)	lifetime of emission (deoxygenated)τ(Ar) (μs)	quantum yield of emission (deoxygenated)Φ(Ar)
Fb1	640	708	0.26	0.0096	0.046
Zn1	599	653	0.16	0.0013	0.013
Ir1	680	755	0.98	90.0	0.022
Pd1	672	739	1.30	748.2	0.039
Pt1b	651	710	1.76	49.6	0.082

^aThe ratio of the integrated intensity of Q_em(0,0) to the integrated intensity of Q_em(0,1) is also presented.

^bData previously published.³⁶

Lifetimes, τ_(Ar) and quantum yields, Φ_(Ar), were measured in argon-saturated solutions and can be split into two categories: fluorescent and phosphorescent porphyrins. Focusing on the fluorescent porphyrins, Fb1 and Zn1, their lifetimes are much shorter, 9.6 and 1.3 ns, respectively, which is typical of fast fluorescence from the S₁ state. Incorporation of the Zn(II) ion promotes intersystem crossing from the S₁ state to the T₁ state through increased spin orbital coupling via the heavy atom effect. Therefore, Zn1 has a much shorter τ_(Ar) and a lower Φ_(Ar). Moving onto the phosphorescent porphyrins, their emission lifetimes are much longer, on the order of μs, indicating slow phosphorescence from the T₁ state. The phosphorescence τ_(Ar) and Φ_(Ar) of Pt1 (τ_(Ar) = 49.6 μs and Φ_(Ar) = 0.082) and Pd1 (τ_(Ar) = 748.2 μs and Φ_(Ar) = 0.039) are governed mainly by the increasing heavy atom effect of the Pt(II) compared with that of the Pd(II) ion, which decreases τ_(Ar) and increases Φ_(Ar). Ir1 by comparison has a smaller Φ_(Ar) = 0.022 than both Pt1 and Pd1 and a τ_(Ar) = 90.0 μs, which is longer than Pt1 but shorter than Pd1.

Polystyrene PSP Performance Studies

The metalloporphyrins were formulated into polymer PSPs, using the polystyrene formulation from our previous luminophore comparison study.³⁶ In this study, however, a higher M_w of polystyrene was used (M_w = 380,000). Polystyrene is not the optimal polymer for PSP formulations but was deemed suitable for luminophore comparison studies. The performance metrics studied are pressure sensitivity, S_p, temperature sensitivity, S_T, and photodegradation after constant illumination at room temperature and pressure for 45 min.

Pressure Sensitivity, S_p

Pressure sensitivity, S_p, is a crucial performance metric for PSPs; a higher sensitivity to pressure allows for smaller pressure changes to be resolved. Stern–Volmer calibrated responses at 273, 293, and 313 K are presented below (Figure 7 and Table 4). S_p is calculated as the change in I_ref/I, with respect to P/P_ref, and is quoted at 293 K (denoted as S_p (293 K)). The phosphorescence lifetime, τ_(Ar), of the luminophore is an important determining factor of S_p; the longer the T₁ excited state exists, the more chances there are for collisional quenching with diffused O₂. There is no significant change in the emission and excitation spectra of the porphyrins in polystyrene. The phosphorescent lifetimes in argon-saturated polystyrene were found to be biexponential for the phosphorescent porphyrins (this is common for luminophores immobilized in a polymer matrix) and monoexponential for the fluorescent porphyrins.⁵³ For the phosphorescent porphyrins, the S_p (293 K) increases in the order of Pt1 = 0.662 < Ir1 = 0.824 < Pd1 = 0.913, which correlates to the trend in the τ_(Ar) of these luminophores in polystyrene. Therefore, Ir1- and Pd1-containing PSPs are much more sensitive to pressure than Pt1-containing PSPs and are attractive options for improved PSP formulations. The small changes in I_ref/I with increasing pressure for Fb1 and Zn1, make accurate performance determination unreliable, and they would not be able to accurately sense large changes in pressure. The fluorescent porphyrins had S_p (293 K) = 0.019 for Fb1 and 0.007 for Zn1. On the fluorescent time scale of nanoseconds these luminophores are very limited by oxygen diffusion through the polystyrene binder. However, the reduced fluorescent lifetime of Zn1 may make it an attractive option for fast responding porous PSPs, where sensing is theoretically not limited by oxygen diffusion through the binder and thus is something we plan to explore in the future.⁵⁴ Overall, the S_p (293 K) of the phosphorescent porphyrin PSPs presented here varies by 38%, highlighting that the nature of the central metal ion can have a substantial effect on S_p.

Figure 7.

(Left) Luminescence response with increasing pressure of Fb1, Zn1, Ir1, Pd1, and Pt1 polystyrene PSPs at 293 K. (Right) Modified Stern–Volmer calibrated luminescent response to pressure with associated linear fits of Fb1, Zn1, Ir1, Pd1, and Pt1 polystyrene PSPs at 293 K.

Table 4. Pressure Sensitivity S_p of Fb1, Zn1, Ir1, Pd1, and Pt1 Polystyrene PSPs at 273, 293, and 313 K with Associated Standard Errors^a.

porphyrin	pressure sensitivity, Sp			weight-averaged emission lifetime in argon-saturated polystyrene ⟨ τ⟩ (μs)
	273 K	293 K	313 K
Fb1	0.015 ± 0.001	0.019 ± 0.001	0.018 ± 0.001	0.0156
Zn1	–0.004 ± 0.001	0.007 ± 0.001	0.007 ± 0.001	0.00157
Ir1	0.564 ± 0.006	0.824 ± 0.005	1.175 ± 0.010	96.8
Pd1	0.624 ± 0.004	0.913 ± 0.009	1.228 ± 0.019	891.4
Pt1	0.448 ± 0.004	0.662 ± 0.005	0.970 ± 0.008	55.3

^aThese values were found from the gradient of the modified Stern–Volmer calibrated plots at each temperature. The weight average lifetimes in argon-saturated polystyrene were calculated using a bi-exponential decay fit and eq 1.

Temperature Sensitivity, S_T

The temperature sensitivity, S_T, of PSPs is also an important performance characteristic. PSPs and polymer PSPs especially possess a high inherent temperature sensitivity, which makes reliable pressure determination more convoluted. The temperature sensitivity of PSPs is due to two factors: at higher pressures the temperature-dependent oxygen diffusion through the binder dominates and at lower pressures the intrinsic temperature dependence of the nonradiative deactivation of the luminophore excited state dominates.¹ A low S_T, single luminophore PSP would be ideal to avoid the complexity associated with binary PSP formulations. The long-lived T₁ excited state has a temperature dependency due to vibrational, rotational, and collisional quenching. Keeping the polymer binder the same, while changing the luminophore, theoretically allows investigation of the luminophore effect on PSP temperature sensitivity. S_T is calculated as the change in I_ref/I with respect to the temperature and is typically quoted at 100 kPa, denoted as S_T (100 kPa). Looking at the phosphorescent porphyrins first, it can be seen that the S_T (100 kPa) increases in the following order: Pt1 = 1.56%/K < Ir1 = 1.62%/K < Pd1 = 1.71%/K, aligning well with their increasing emission lifetimes. The change in S_T with increasing pressures is also presented for Ir1, Pd1, and Pt1 (Figure 8). For Ir1 and Pd1, S_T (10 kPa) ≈ 0.20%/K; for Pt1, it is slightly higher, with S_T (10 kPa) = 0.32%/K. As the pressure increases, S_T increases linearly for Ir1 and Pt1. However, the increase is greater for Ir1 than for Pt1, causing the S_T of Ir1 to become greater than that of Pt1 at roughly 65 kPa. At higher pressures, there is more oxygen in the binder matrix, and the longer lived T₁ lifetime of Ir1, compared to Pt1, will facilitate more deactivation by this increased O₂ concentration. Therefore, Ir1 at these higher pressures will be more susceptible to the increasing collisions with O₂ because of the increasing energy of diffused O₂, which accompanies increasing temperatures. Pd1 has an even longer T₁ lifetime than Ir1 and so its S_T becomes greater than that of Pt1 earlier (roughly 40 kPa). However, in contrast to Ir1 and Pt1, for Pd1, a second-order polynomial fit was found to better fit the change in S_T with pressure, compared to a linear fit (R² = 0.999 for second-order polynomial and 0.991 for linear). This nonlinear behavior is probably due to excess quenching of Pd1 at higher pressures and consequently causes the S_T of Pd1 to plateau and subsequently dip below that of Ir1 at pressures >130 kPa. These findings are interesting, as they implicate that the more pressure-sensitive Ir1 and Pd1 PSPs could be used in low-pressure (<70 kPa) measurements where their S_T is lower and their low quantum yields (at high pressures) are less of an issue. On the other hand, Pt1-based PSPs are more suited to high pressure (>70 kPa) applications, where the S_T is lower (compared to Ir1 and Pt1) and the higher quantum yields are more desirable. The two fluorescent porphyrins, Fb1 and Zn1, exhibit weak temperature sensitivities, 0.13 and 0.11%/K, respectively, due to their much shorter fluorescent lifetimes.

Figure 8.

Change in temperature sensitivity, S_T, of Ir1, Pd1, and Pt1 polystyrene PSPs with increasing pressure. These values were found from the change in I_ref/I across the three temperatures of the modified Stern–Volmer calibrated plots. The linear fits (Ir1 and Pt1) and the second-order polynomial fit (Pd1) of this data are also included.

Photodegradation

Photodegradation of the phosphorescent porphyrins Ir1, Pd1, and Pt1 polystyrene PSPs was also investigated. The photodegradation is calculated as the % luminescent intensity of the original luminescent intensity (at 0 min) after 45 min of constant illumination at room temperature and pressure. A high-performing PSP formulation will have minimal photodegradation during testing to facilitate greater usage over multiple tests. Photodegradation of PSPs during testing occurs primarily via the formed singlet oxygen radical, which can go on to damage the luminophore and binder, leading to a reduction in paint performance.⁵⁵Pt1 photodegrades the most after 45 min, with a 6.6% reduction in luminescence intensity (Figure 9). Pd1 is the most photostable, with a 4% reduction in luminescence intensity after 45 min of constant illumination. Ir1 is more photostable than Pt1, with a 5.5% photodegradation after 45 min. However, the photodegradation of Ir1 is noticeably more linear in this time scale (compared to Pt1), which begins to tail off toward the end of the measurement.

Figure 9.

Photodegradation of Ir1, Pd1, and Pt1 polystyrene PSPs, at room temperature and pressure, over 45 min of constant illumination. I_ref is the luminescent intensity at 0 min.

Conclusions

We have investigated a family of metalloporphyrins and their parent free-base ligand TFPP to examine the effect of the central metal ion of porphyrins on the electronic structure, photophysics, and the resulting polymer-based PSP performance. The Pt(II) complex, PtTFPP (Pt1 in this study), is used almost ubiquitously in polymer-based PSP formulations and therefore we presented Pd(II) (Pd1), Zn(II) (Zn1), and, for the first time in the PSP field, Ir(III) (Ir1) TFPP complexes. The central metal ion mainly affects the energy of the occupied a_2u MO through metal electronegativity-dependent stabilization. Preferential stabilization of the a_2u MO by more electronegative metals widened the HOMOs to LUMOs energy gap, resulting in a blueshift of the spectra. However, Ir1 was an exception, as it was found using DFT calculations to have stabilized LUMOs relative to the other metalloporphyrins, resulting in red-shifted spectral features. The preferential stabilization of the a_2u away from the a_1u MO increases the HOMO to HOMO–1 energy gap and thus increases the intensity of Q(0,0) feature in the absorption and emission spectra. Ir1 was found to have a greatly stabilized a_1u MO, brining it closer in energy to the a_2u MO, thus reducing the intensity of the Q(0,0) features. TD-DFT calculations showed that the intensity of the Q_abs(0,0) feature increased due to the contributions of the electronic transitions from the HOMO–1 and the HOMO to the LUMOs, in the excited-state makeup decreasing and increasing, respectively, with increasing HOMO–1 to HOMO energy gaps. The heavy metal effect of the Zn(II) ion caused Zn1 to have a reduced fluorescence lifetime and quantum yield, compared to the freebase porphyrin Fb1. On the other hand, Pd1, Ir1, and Pt1 exhibited room-temperature phosphorescence, with Pt(II) having a larger quantum yield (0.039 and 0.082, respectively) but shorter lifetime of emission (748.2 and 49.6 μs, respectively) due to the increased heavy atom effect of Pt(II) over Pd(II). Ir1 had a smaller quantum yield (0.022) than Pd1 but a longer lifetime of emission (90.0 μs) compared to that of Pt1. The phosphorescent porphyrins Ir1, Pd1, and Pt1, were all viable candidates as luminophores in PSP formulations. The pressure sensitivity at 293 K, S_p (293 K), of the resulting polystyrene PSPs greatly increased in the order of Pt1 = 0.662 < Ir1 = 0.824 < Pd1 = 0.913, in line with the increasing phosphorescence lifetimes. Increasing the phosphorescence lifetime of the luminophore was found to increase the temperature sensitivity (S_T) at higher pressures, with Ir1 and Pd1 having higher S_T at higher pressures than Pt1. However, at lower pressures, Pt1 had the highest S_T and Ir1/Pd1 had a lower S_T. Therefore, the nature of the central metal ion in porphyrin luminophores can be used to tailor PSP formulations for specific pressure measurement. The quantum yields of emission for Pd1 and Ir1 are also much lower than for Pt1, reducing the brightness of these PSPs. Finally, this work shows that the central metal ion of porphyrin luminophores can be used to tune the pressure and temperature sensitivity of the resulting PSP formulation for the desired application. This work contrasts our previous study on the effect of porphyrin phenyl halogenation, which mainly affected the resulting temperature sensitivity of the PSPs.³⁶ We hope this research can assist with making new and improved PSP formulations.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.4c09045.

• Details and spectra of syntheses and characterization of compounds and their NMR, mass spectrometry and IR data, crystallographic data, DFT calculations, luminescence data, and polystyrene PSP performance data (PDF)

• Crystallographic data for Zn(II) TFPP complex (CIF)

• Crystallographic data for Ir(III) TFPP complex (CIF)

• Crystallographic data for Pd(II) TFPP complex (CIF)

The authors declare no competing financial interest.