Introduction

Over the past decade, the field of synthesizing new organic compounds with tailored photochemical and/or photophysical properties has attracted much attention due to their fertility and the possibility of exploiting their properties in many potential optoelectronic applications. Today, the highly cost and fragile inorganic electronics have been replaced by their flexible, and affordable organic molecular counterparts, which possess high optical absorption coefficients and promising electrical properties. Undoubtedly, the recent studies on these organic compounds, specially phthalocyanines and porphyrins and their metal complexes, have drawn attention to the importance of the rigorous structure–property relationship of the compound, which has led to noticeable progress in the efficiencies of the optoelectronic devices [1, 2].

Functionalization is one of the most important approaches to controlling the properties of these compounds by adding a functional group in a peripheral position in the stereoscopic molecular space or by introducing a metal ion (whether transitional metal or rare earth element) to the compound. It is worth mentioning that the metal ion plays an effective and main role in controlling the optoelectronic characteristics of a material by controlling its orientation and molecular coordination, giving rise to unique molecular design, improving the mobility of its charge carriers, and reducing the dynamical molecular rigidity, which in turn increases the possibilities of radiative transitions [1, 2].

Interestingly, the high conjugation and the modulation’s flexibility of the optoelectronic properties of porphyrins made these macrocyclic compounds a versatile research platform and be intensively exploited in many applications. Based on the type of substitution, either donating or acceptor group, and according to the position at which the substitution is introduced, either meso and/or beta, the molecular packing, optoelectronic and electrochemical properties of the planar porphyrin compound can be systematically determined [1,2,3,4]. Furthermore, the molecular geometry conformation of the metalloporphyrin can be varied according to the relative alignment of the ligand-central metal planes as well as the metal’s size giving rise to either in-plane structure (metal size 55–80 pm) or out of the plane structure (metal size 80–90 pm) such in case of FeTPPCl and MnTPPCl [4].

Metalloporphyrins are well-known compounds for their unique optical absorption properties by their characteristic Soret and Q-bands as well as optoelectronics owing to the existence of multiple electronic and spin states. Moreover, they were used as analytical reagents for separating different metal ions owing to their high stability and high UV–visible absorption characteristics [4, 5]. On the biological and biomedical level, metalloporphyrins are considered one of the most important and most widely used compounds not only in bioimaging and drug delivery but also as active material in different architectural designs of biosensors to detect different compounds such as toxic, vitamins, biogenic, polyphenolic and amines compounds from pharmaceutical, biological samples, and food products [4, 6]. Furthermore, the NIR absorption features of metalloporphyrins were used for cancer photodynamic therapy [1]. Furthermore, since metalloporphyrins have distinctive chemical features such as catalytic activation, optical changes, and reversible binding, they are intensively assigned as an active layer in chemosensors such as gas sensors [7], PH sensors, explosive gasses sensors, and volatile organic compounds sensors [2].

On the other hand, metalloporphyrins, specifically transition metal tetraphenyl porphyrins, achieved high performance in organic solar cell technology. Metalloporphyrins were used in different designs such as dye-sensitized solar cells with a maximum PCE ~ 12.3% [8,9,10,11], organic/inorganic heterojunction solar cells with PCE in the range of (2–5.6) % [12,13,14], hybrid molecular solar cells [15], and bulk heterojunction solar cells [16]. Moreover, metalloporphyrins are employed as the main constituent of organic light-emitting diodes (OLEDs) manufacture [17, 18] and broadband photodetectors [19].

Therefore, in light of the intensified interest in the effect of the central metal ion in the composition of tetraphenylporphyrin on the optoelectronic properties [20], we have studied the behavior of some tetraphenylporphyrin compounds such as copper tetraphenyl porphyrin (CuTPP), nickel tetraphenyl porphyrin (NiTPP), iron tetraphenyl porphyrin chloride (FeTPPCl), and manganese tetraphenyl porphyrin chloride (MnTPPCl) in the form of an organic/inorganic heterojunction as a photodetector and interpreting the results in light of the structural, molecular and optical properties of the grown tetraphenylporphyrin films.

Materials and methods

Thin films fabrication

5,10,15,20-Tetraphenyl-21H,23H-porphine copper (II) [C44H28CuN4], 5,10,15,20-tetraphenyl-21H,23H-porphine nickel (II) [C44H28NiN4], 5,10,15,20-tetraphenyl-21H,23H-porphine iron (III) chloride [C44H28ClFeN4], and 5,10,15,20-tetraphenyl-21H,23H-porphine manganese (III) chloride [C44H28ClMnN4] of molecular weight 676.27, 671.41, 704.02, and 703.11 g/mol, respectively, are the target metalloporphyrins and were purchased from Sigma-Aldrich. The molecular structures of the metalloporphyrins compounds under study are shown in Fig. 1a. Thin films of the purchased metalloporphyrins without the need for any further purification were fabricated under a vacuum of pressure (2 × 10−5 mbar) using Edwards 306A thermal evaporator coating unit onto well-cleaned quartz substrates. During the deposition process of the four porphyrins, the evaporation rate, film thickness, and all the deposition parameters are kept constant to exclude their influence on the deposited films’ properties. The distance between the rotatable substrates holder and the well-outgassed quartz crucible heated by the molybdenum boat was fixed at about 21 cm to ensure film thickness uniformity. The obtained film thickness and evaporation rate were monitored in situ by the Edward FTM6 crystal monitor and found to be approximately 181 ± 3 nm and 0.2 nm/s, respectively.

Figure 1
figure 1

a Chemical structure of MTPP [M = Cu(II), Ni(II), Fe(III)Cl, Mn(III)Cl], and structure and measurement setup

Full size image

Device fabrication

For photodetector fabrication, the metalloporphyrin thin films were deposited on the surface of chemically etched n-type silicon substates of orientation < 100 > . Next, a thin layer of lithium fluoride, LiF, of thickness 9 nm was deposited on the surface of metalloporphyrin as a buffer layer to protect the organic layer from diffusing the metal atoms of the top electrode into it. The heterojunction's electrodes were then deposited, with magnesium (55 nm) and aluminum (183 nm) acting as the top and rear electrodes, respectively, on the LiF/MTPP/n-Si, resulting in the Mg/LiF/MTPP/n-Si/Al architecture seen in Fig. 1b. Herein, a low work function metal electrode is utilized as a top electrode to facilitate the charge carrier’s transportation from metal and organic molecular orbitals.

Characterizations instrumentation

The crystal structure of the deposited metalloporphyrins thin films was inspected using X’pert pro-Panalytical, X-ray diffractometer in the grazing mode for thin films in the range of (5–80) degree with an X-ray beam of a wavelength of about 1.5418 Å. Furthermore, the stability of MTPP’s molecular structures after deposition is evaluated with the aid of the confocal Raman microscope model [Witec 300R alpha, Germany] with excitation wavelength 532 nm and laser power of about 2 mW.

The prepared metalloporphyrin films' topographical features were inspected using WITec atomic force microscope in contact mode. The X-ray photoelectron spectroscopy (XPS) analysis was performed on K-ALPHA (Thermo Fisher Scientific, USA) with monochromatic X-ray using Al Kα radiation as an exciting source of energy (1486.6 eV) and spot size 400 μm at pressure 1 × 10−9 mbar with a full-spectrum pass energy 200 eV and narrow-spectrum 50 eV. The core level XPS characteristic features of C1s, N1s, Cl2p, and M2p [M = Cu, Ni, Fe, Mn] are probed and properly deconvoluted after subtracting the Shirley-type background and using Voigt line shape function. The XPS signals of the respective elements of the prepared metalloporphyrins samples are corrected by their proper sensitivity factor. The full width at half maximum (FWHM), the area under peak (area), peak height, binding energy (BE), and atomic percent of the resolved core level peaks for the four metalloporphyrins are listed in table S1 (supporting information).

Moreover, the optical features of the deposited MTPP films in the spectral range from UV to NIR were evaluated based on the measured absorption, Abs (λ), reflection, R(λ), and transmission, T(λ), spectra in the range of (200–2500) nm at room temperature. The measurements were taken using JASCO 570 double beam spectrophotometer at normal incidence. In addition, the photoluminescence spectra, PL (λ), of the fabricated films were measured using a spectrofluorophotometer [model: Shimadzu RF‐ 5301PC]. Finally, the designed photodetector's light-sensing properties were investigated using Keithley electrometer 6517B at forward and reverse bias conditions in swee** voltage range ± 5 V under dark and halogen lamp illumination in Fig. 1b. In addition, the intensity of the incident light was controlled by IL1400A radiometer/ photometer to be varied from 20 to 100 mW/cm2.

Results and discussion

XRD crystal structure analysis

The XRD patterns of the deposited CuTPP, NiTPP, FeTPPCl, and MnTPPCl thin films depicted in Fig. 2 declare the partial crystallization of the grown films with a dominance of amorphous background owing to the existence of weak diffraction peaks with two dominant broad humps at position 7.91 ± 0.12 nm and 20.76 ± 0.3 nm. This confirms the short-range order with a noticeable amorphous disorder. Furthermore, the estimated crystallite size for the indexed weak diffraction peaks were about 9.3, 11.3, 10.4, and 11.6 nm for CuTPP, NiTPP, FeTPPCl, and MnTPPCl, respectively. Consequently, it can be noted that the grown films are composed of a nanostructured crystallite distributed in an orderly manner for a short range in the presence of the dominance of the amorphous nature. These observations about the randomness of the crystal structure of the thermally grown metalloporphyrins thin films were previously reported in much research [21,22,23,24].

Figure 2
figure 2

XRD diffractogram of a CuTPP, b NiTPP, c FeTPPCl, and d MnTPPCl as-deposited thin films

Full size image

Raman molecular structure analysis

The power of Raman spectroscopy is exploited here to identify the vibrational dynamics of the molecular skeleton and the functional groups attached to the compounds under study. Figure 3 reveals the Raman spectra of CuTPP, NiTPP, FeTPPCl, and MnTPPCl thin films in the wavenumber range (100–2000) cm−1. Most of the Raman active bands, prominent in the figure, are grouped and listed in Table 1 with their corresponding assignments. The characteristic peak of the in-plane transitional motion of phenyl groups within the porphyrin is observed at 203–208 cm−1. The Raman active mode of metal-N stretching vibration is observed for the four compounds in the band of (391–405) cm−1 [25, 27, 29]. These shifts in the position of this sensitive mode are an account of the difference in C-N and Metal-N bond length for each compound [36]. The M–N vibrational band may be overlapped with the contraction/expansion mode of the porphyrin ligand [30]. The characteristic band of M-Cl [M = Fe, Mn] stretching vibrations is detected at 363 and 365 cm−1 for Fe-Cl and Mn-Cl, respectively [27,28,29, 36]. The intensity of these two peaks is very weak, which may be related to the decomposition of the M-Cl [M = Fe, Mn] bond during the thermal evaporation of the films. These two structures, i.e., FeTPPCl and MnTPPCl, are nonplanar structures owing to exceeding ionic radii 90 pm [4, 36]. Furthermore, the featured Raman band corresponding to the aromatic pyrrole and phenyl ring breathing in-phase vibrations and the stretching C–C vibrations are noticed in the four compounds' spectra at around 1005 cm−1 [25, 29, 30, 34]. The bands in the range of (1228–1238) cm−1 are ascribed to the stretching vibrations of C–C phenyl [25, 29, 30, 34]. The symmetric pyrrole half-ring stretching vibrations overlapped with the asymmetric C-N stretching vibrations are noticed at (1362–1375) cm−1; meanwhile, this large variance may be due to the slight involvement of N–H in-plane bending [25, 26, 29, 30, 32, 34]. The Raman shoulder at 1585 cm−1 emphasizes the C–C stretching vibration of the phenyl ring [25, 30].

Figure 3
figure 3

Raman spectra of a CuTPP, b NiTPP, c FeTPPCl, and d MnTPPCl as-deposited thin films

Full size image
Table 1 Raman active bands for the MTPP [M = Cu, Ni, FeCl, MnCl] thin films
Full size table

However, the metalloporphyrins exhibit featured aromatic character; their molecular structure may undergo a variety of intrinsic nonplanar distortions such as ruffling and saddling [30, 31, 34, 37]. The nonplanarity deviations can significantly influence the metalloporphyrins' photochemical and photophysical properties due to the strong interaction between porphyrins and phenyls π-systems, yielding a splitting of occupied molecular orbitals as well as redshift and broadening of the optical absorption and emission spectra, thereby increasing the exciton binding energy [38,39,40]. It is heartening to note that the CuTPP core does not undergo significant out-of-plane distortion [30], but in contrast, NiTPP tends to deviate from planarity via the saddling and ruffling distortions owing to the shortened Ni–N bond (1.941 Å) [34]. Nonplanar distortion in NiTPP thin film is identified by the presence of two Raman bands at 330 and 801 cm−1 related to saddling and ruffling, respectively [25, 34].

AFM morphological characterizations

The surface manifestations of the deposited metalloporphyrins thin films are inspected in terms of the 3D AFM profiles of two different areas 50 × 50 μm2 and 5 × 5 μm2 shown in Fig. 4a–d. Homogenous nanostructured films of rough herringbone surfaces are revealed. The surface of prepared thin films seems to be continuous with randomly oriented and loosely packed domains of RMS roughness of about 25.9, 50.5, 16.86, and 43.78 nm for CuTPP, NiTPP, FeTPPCl, and MnTPPCl thin films, respectively. This high value of roughness would significantly improve the performance of the heterojunction since the absorption cross section of the fabricated device would be enhanced through the light-harvesting [14, 41]. The robustness of the film's surface may be attributed to the significant role of the metal ion–ligand bonding [64]. The area ratio between the M2p3/2 and M2p1/2 components is approximately 2, which endorses the multiplicity of the degenerate M2p3/2 and M2p1/2 electronic distribution.

Figure 7
figure 7

Resolved core level XPS spectra of a Cu2P, b Ni2P, c Fe2P, d Mn2P, e Cl2p of FeTPPCl, and f Cl2p of MnTPPCl

Full size image

The well-screened peak of 2 + oxidation state of Cu2p3/2 at 934.34 eV is attributed to the 2p3d10L final state (2p represents a core hole state and L a ligand hole state), and the shake-up satellite peaks positioned at 940.81, 942.64, and 944.40 eV are owing to the photoelectrons from the 2p3d9 final state [44, 65]. The XPS shoulder at 932.87 eV observed in Fig. 7a may correspond to the 2p3d10 final state, ensuring Cu has a 3d10 ground state [65]. On the other hand, the main formal 2 + oxidation state of Ni2p3/2 at 854.59 eV is an accountant to the 2p3d9L final state, and the two satellite peaks at 3.5 and 7.76 eV away from the main Ni2p3/2 photopeak in the higher binding energy direction owing to the 2p3d8L final state [66]. Additionally, a single satellite peak associated with Ni2p1/2 photopeak away by 8.23 eV is observed as detected in Fig. 7b.

Furthermore, the highly resolved Fe2p XPS spectrum that is revealed in Fig. 7c yields two bands at (705–715) eV and (720–732) eV that are related to 2p3/2 and 2p1/2 with their corresponding satellites. The first band has two peaks at 707.85 and 709.94 eV related to Fe2+ 2p3/2 and Fe3+ 2p3/2, respectively, and a satellite at 714.39 eV. Meanwhile, the second band contains two photopeaks at 721 and 723 eV that are ascribed to Fe2+ 2p1/2 and Fe3+ 2p1/2, respectively, and a satellite at 729.09 eV [49, 62, 67, 68]. Interestingly, the positively Feδ+ is balanced by nitrogen in the Fe–N consistent with the observed decrement of Fe–N contribution to N1s XPS signal in Fig. 6g, which is analog to the existence of two oxidation states in Fe2p [56, 69, 70]. Likewise, the resolved Mn2p XPS signal revealed in Fig. 7d unveils two resolved peaks at 641.7 and 653.28 eV related to Mn3+ 2p3/2 and Mn3+ 2p1/2, respectively, with their corresponding satellites at 644.54 and 659.54. The contribution of Mn2+ cannot be detected in the present XPS maybe due to the oxidation of the film’s surface [51, 56]. The multiple structures of Fe and Mn may be ascribed to their open shell structure where coupling between core hole spins and valence shell spins, resulting in multiple final states taking place [51, 56]. The Cl2p spectra either of FeTPPCl and MnTPPCl shown in Fig. 7e, f, respectively, are resolved and result in two distinct peaks at 198.2 and 200.8 eV ascribed to 2p3/2 and 2p1/2, respectively [51]. This indicates that the single chemical state of the remained Cl is responsible for the axial coordination with either Fe or Mn [51].

Optical characterizations

Transmission and reflection spectra

Figure 8a–d reveals the transmission, T(λ), and reflection. R(λ) spectra profiles for the utilized metalloporphyrins of the same film thickness over a wide spectral range extending from UV (190 nm) to NIR (2500 nm). From the following spectra, it can be concluded that the deposited films behave as transparent films (90% < T(λ) < 100%) in the NIR spectral range (λ > 900 nm) and possess high absorption in the UV and visible regions. As frequently, the fundamental absorption edge is observed at 468, 444, 438 and 490 for CuTPP, NiTPP, FeTPPCl, and MnTPPCl, respectively [71,72,73,74]. Within the transparent region, the spectral behavior of MTPP [Cu, Ni] and MTPPCl [Fe, Mn] is similar. However, the featured absorption peaks of metalloporphyrins in the absorbing region are sufficiently varied owing to the metal ion incorporated inside the porphyrin ring.

Figure 8
figure 8

Transmission and reflection spectra of a CuTPP, b NiTPP, c FeTPPCl, and d MnTPPCl as-deposited thin films

Full size image

Absorption spectra

The typical absorption spectrum of metalloporphyrins depends mainly on the π–π* electronic transitions between the two HOMOs [a1u(π), a2u(π)] and the two LUMOs [twofold degenerate eg(π*)] of the conjugated porphyrin ring [43, 71, 72, 75]. These molecular orbitals can be perturbed by the central metal ion according to Gouterman’s four-orbital model. The absorption coefficient, α(λ), spectra of the deposited metalloporphyrins described in Fig. 9a–d cover a wide range of wavelengths extending from ultraviolet to visible regimes, which suggests these materials for efficient wide-band photodetection applications. The absorption spectra of metalloporphyrins depict three main bands that are well known for porphyrins; the pronounced intense Soret band (B-band) in the spectral range (330–500) nm that arises from highly allowed (S0 → S2) electronic transitions, the faint Q-band in the spectral range (500–700) nm that arises from weakly allowed (S0 → S1) electronic transitions in addition to the (C, L, M, N) bands in the UV region [76]. Further inspection of the spectral distributions of α(λ) of the present metalloporphyrins is performed and resulted in a set of highly resolved Gaussian oscillators to fit the B and Q-bands, where the energetic positions of each oscillator used in fitting are listed in Table 3. The peaks’ positions of the four metalloporphyrins understudy remarkably agrees to that were reported in previous work [71, 72, 74].

Figure 9
figure 9

Absorption coefficient spectra of a CuTPP, b NiTPP, c FeTPPCl, and d MnTPPCl as-deposited thin films

Full size image
Table 3 Optical parameters of the fabricated metalloporphyrin thin films
Full size table

Interestingly, the UV absorption bands (C, L, M, N) elucidate the charge transfer process from the metal d-orbitals to the ligand’s low-lying π* molecular orbitals. The sensitivity of such weak absorption bands to the variation of central metal ion of the porphyrin cavity is lower than that of Q and B bands [20]. The C-band located at 6.31 ± 0.03 eV is ascribed to the dπ* transitions of symmetry b1g that originates from the partially filled d-orbitals (dx2-y2) of the transition metal to the porphyrin molecular orbitals [25, 44, 61, 72]. Likewise, the N-band positions are about 4.33, 4.43, 4.46, and 4.5 eV for CuTPP, NiTPP, FeTPPCl, and MnTPPCl, respectively. This shift is directly correlated with the elevation of dπ levels responsible for this transition as one moves from Cu to Mn in the periodic table [38, 72]. The N and M-bands are ascribed to the electronic transitions from HOMO level 3a2u(π) and 1b1u (π) to 5eg(π*) LUMO, respectively [38, 76, 77]. Next, the observed intense Soret B-band and weak Q-band result from the electronic transitions from to a2u (π) HOMO to eg (π*) LUMO and from a1u (π) HOMO to eg (π*) LUMO, respectively [19, 38, 74, 76]. Particularly, the intensity ratio between Q and B bands measures the mixing degree of transition metal d-orbitals and porphyrin molecular orbitals, where the greater the mixing degree, the closer the energy of unmixed states [20]. The estimated values of Q/B are about 0.29, 0.39, 0.19, and 0.15 for CuTPP, NiTPP, FeTPPCl, and MnTPPCl, respectively, which coincident with the elevation of d-orbitals with moving from Cu to Mn in the periodic table [20, 38]. According to Davydov splitting hypothesis [78], the interaction between two or more molecules in a single unit aggregate cell may cause the splitting of the absorption band into two or more excitonic transition peaks due to the excitation states vibronic coupling. The Davydov spectral splitting is ruled by the density and inter-distance of the interacting molecules in addition to the localized transition dipole moment orientation, where the B-band’s oscillator strength is higher than that of the Q-band due to the in-phased dipole moments [71,72,73,74,75,76]. This is observed in the splitting of the B-band into a set of spectral peaks with different energies (BX1, BX2, BY1, and BY2) as predicted in the theoretical calculations in the literature [38, 71, 72, 74, 79], particularly in the case of MTPPCl [M = Fe, Mn] owing to the large ionic radii of the inserted central cavity metal ion FeCl and MnCl compared to Cu and Ni ions. The obtained values of Davydov splitting in B-band are about 0.15, 0.35, 0.43, and 0.71 eV for CuTPP, NiTPP, FeTPPCl, and MnTPPCl, respectively.

Furthermore, the features of Q -band change quite dramatically upon changing the central metal ion in the porphyrin ring. It contains two peaks (Qα and Qβ) as shown in the insets of Fig. 9a–d, which are ascribed to pure (S0 → S1) electronic transitions with no contribution of molecular vibrations (Qβ) and a transition from the ground state (S0) to higher vibrational levels in S1 (Qα) [76, 80]. The relative intensity of Q-band peaks (Qα/Qβ > 1) confirms the good stabilization of metal ions with porphyrin ligand in a square planar structure in the case of CuTPP, NiTPP, and FeTPPCl. Against by, MnTPPCl exhibits (Qα/Qβ < 1) with a shoulder at 1.98 eV confirming the less stability of the complex and deviation from planarity [76, 80]. It should be noted that the intensity of the B-band and the UV absorption bands of CuTPP film is higher than those of other metalloporphyrins in addition to conserving its planarity, which may increase its applicability for more efficient photodetection.

Energy gap calculation

Moreover, the indirect energy gap, Eg, and the onset energy, Eonset, are estimated utilizing Tauc’s method, as represented in Fig. S1 (supporting information). The resulting values of energy gap and onset energy are in good agreement with those recorded in previous work [38, 71,72,73,74, 77, 79]. From the obtained results, CuTPP exhibits the lowest energy gap (HOMO–LUMO gap) compared to the other analogs, which in turn renders the photoexcitation process more facile, yielding a higher density of photogenerated charge carriers. The low energy gap of CuTPP film may be a consequence of the π–π molecularly interacted arrangement of CuTPP molecules owing to their unique planarity [81]. Nonetheless, CuTPP also records a relatively high value of Eonset compared to other metalloporphyrins in the study, which explains that the generation of bounded Frenkel exciton in it may require more energy than other metalloporphyrins counterparts, but the transition process requires less energy. Furthermore, the wider HOMO–LUMO gap of the metalloporphyrin molecule is related to the more chemical hardness, which means that the electronic cloud distribution would impede the deformation that arises from the electron transfer from the occupied orbital of one molecule to the empty orbital of the neighboring one [81].

Photoluminescence spectra

Figure 10 illustrates the obtained photoluminescence spectra of the fabricated metalloporphyrins thin films. Upon photoexcitation at 350 nm, the second sing-doublet excited state (2S2) is populated, which relaxes to the S0 ground state giving rise to a sharp Soret band [82, 83] that is observed at 397.9 nm and 384.8 nm accompanied by a shoulder at 334.4 nm and 341.9 nm for CuTPP and NiTPP, respectively. On the other side, MTPPCl [M = Fe and Mn] shows notably split PL spectra at wavelengths 316.2 nm and 391.4 nm for Fe and 360.2 nm and 430.5 nm for Mn. The photoluminescence of FeTPPCl and MnTPPCl in the split form is less efficient than that of CuTPP and NiTPP, which agrees with the extremely low luminescence quantum yield estimated in previous work [84, 85]. Alternatively, the diminution of FeTPPCl and MnTPPCl PL may be attributed to the rapid intersystem crossing of the sing-doublet state (of very small lifetime < 10 ps) to intermediating state arises from the coupling of trip-quartet states (4T1) and low-lying charge transfer (2CT1) states non-radiatively rather than direct transition to S0 [82,83,84,85,86,87]. Additionally, the presence of Fe2+ and Mn2+ that oxidizes easily may be the reason for the PL diminution. A weak PL peak at 574.9 and 569.1 nm is detected for CuTPP and MnTPPCl, respectively, which may be ascribed to the 2CT2 → S0 transition [83, 86]. The inset of Fig. 10 depicts the zoom-in of the PL spectra in the low energy range, which reveals relatively weak broadband around 670–805 nm with a detectable peak at ~ 751.4 nm for all the studied samples which are ascribed to triblet → singlet ground state transition [82, 86]. It is worth mentioning that the closed electronic shell diamagnetic metalloporphyrins (NiTPP) exhibit strong luminescence compared to the open electronic shell paramagnetic metalloporphyrins (CuTPP), as observed from the inset of Fig. 10. This is explicated in terms of the difference in relaxation pathway in the case of closed-shell metalloporphyrins (2S2 → 2S1 internal conversion) and open-shell metalloporphyrins (2S2 → 2CT1 → 2T1 → 2S0) [82, 86].

Figure 10
figure 10

PL spectra of a CuTPP, b NiTPP, c FeTPPCl, and d MnTPPCl as-deposited thin films

Full size image

Furthermore, the normalized PL profiles are deconvoluted into several Gaussian function subbands as shown in Fig. S2 (supporting information) and compared to the normalized absorption spectra for estimating Stokes shift (absorption–emission difference) in each case. This shift may be ascribed to the EB's exciton binding energy in metalloporphyrin films [88]. It is heartening to note that CuTPP records the minimum value of EB (~ 0.25 eV) compared to the other metalloporphyrins under investigation and other organic semiconductors such as CuPc, α-6 T, NPD, TPD, PTCDA, and Alq3 [88, 89]. The minimum exciton binding energy, the more facile electron–hole separation (exciton dissociation). Therefore, it seems plausible that CuTPP would exhibit superior optoelectronic performance compared to other metalloporphyrins under investigation.

Photodetector characterizations

The photodetection performance of the fabricated vertical organic/inorganic heterojunctions based on metalloporphyrin/n-Si is evaluated in terms of the semilogarithmic relation between current density, J, and applied bias potential, V under dark and different illumination intensities as shown in Fig. 11a–d. Obviously, the current densities of fabricated Mg/LiF/MTPP/n-Si/Al devices increase linearly in the low-bias region and then become quasi-static with voltage independence behavior which is a clear indicator of the good passivation of the n-Si surface and reflects the quality of implemented heterojunctions [90]. The deviation of the J–V semilogarithmic relation from the linearity may be due to the high parasitic series resistance of the fabricated devices. The dark parasitic series resistance of the fabricated devices is estimated using modified Norde’s method [91, 92], as shown in Fig. S3 (supporting information). Notably, CuTPP/n-Si achieves the lowest value of series resistance compared to the other heterojunctions, which may be reflected in its priority and high quality as a photodetector. As a whole, all the fabricated heterojunction devices exhibit a rectification behavior with a noticeable sensitivity toward the incident light intensity's variation, which is conclusive evidence for the eventuality of utilizing them for photodetection applications.

Figure 11
figure 11

lnJ versus voltage of a CuTPP/n-Si, b NiTPP/n-Si, c FeTPPCl/n-Si, and d MnTPPCl/n-Si heterojunctions under dark and illumination conditions

Full size image

The microelectronic parameters, including ideality factor, n, reverse saturation current, Is, open-circuit barrier height, ΦB0, of such devices are estimated in the light of Richardson thermionic emission, TE, theory [90, 92, 93] and listed in Table 4. As can be noticed, all the fabricated devices deviate from the ideal performance (n ≠ 1), which may be attributed to exciton annihilation at the organic/inorganic interface within the depletion region, image force effect, non-homogeneity of potential barrier height, and tunneling effect of the transported charge carriers [90, 92]. In comparison, CuTPP/n-Si heterojunction performance is close to the ideal behavior, while the barrier height is almost equal for all the engineered photodetectors rather than MnTPPCl/n-Si.

Table 4 Electrical parameters of the fabricated metalloporphyrin organic/inorganic heterojunctions
Full size table

The monotonic increase of the heterojunction’s current density upon increasing the illumination intensity (photon flux density) as shown in Fig. 12a is a result of the increased photogenerated electron–hole (e–h) pairs [90, 94]. The fabricated devices show superlinear photocurrent density with exponent (γ > 1), where CuTPP/n-Si and NiTPP/n-Si devices achieve γ ~ 1.6, which declares the existence of localized trap states that are continuously distributed which gives rise to (e–h) recombination [95]. These states may be attributed to the semiconductor un-depleted region’s resistance, the resistance of the oxide layer and dangling bonds at the interface, the resistance of metal-contact in addition to the intrinsic surface, grain, and grain boundary resistances of metalloporphyrin film [90, 92]. The lowest density of these interface trap states is observed in the case of FeTPPCl/n-Si from the profile of interface states density, Nss, distribution with respect to the energy difference, Ec-Ess between the conduction band and trap states energy that is declared in Fig. S4 (supporting information). This follows the low value of reverse saturation current density, Jd, of FeTPPCl/n-Si compared to other metalloporphyrins heterojunctions [96]. The observed Nss values are in the range of 1014 eV−1 cm−2, and it can be noticed that the density of the interface states decreases moderately as their distribution moves far away from the conduction band [90].

Figure 12
figure 12

a lnJ versus lnPin of the fabricated metalloporphyrins/n-Si heterojunction at − 2.5 V, b band diagram of the engineered photodetectors, and c the band offset between n-Si and the utilized metalloporphyrins

Full size image

Intriguingly, the charge carrier transport mechanism in the present photodetection devices can be explained in the light of the purposed band diagram and the band alignment between metalloporphyrins’ frontier molecular orbitals, silicon valance, and conduction bands, and the utilized electrodes Fermi levels as shown in Fig. 12b. Upon illuminating the heterojunction with a halogen lamp light and under the influence of reverse biasing, the electrons of MTPP [M = Cu, Ni] and MTPPCl [Fe, Mn] films would be excited, generating bounded photogenerated e–h pairs at the junction interface, which would be dissociated and drift by the junction’s built-in potential toward the electrodes. The band offset between metalloporphyrins' molecular orbitals (HOMO and LUMO), and n-type silicon's bands (valance and conduction bands) is the main controller of the electrons and holes barriers, as shown in Fig. 12c. As observed, CuTPP/n-Si achieves the lower barrier for holes; consequently, it achieves the best performance compared to other metalloporphyrins heterojunctions under study [90, 92]. The goal of insertion a thick LiF interlayer of wide bandgap as shown in Fig. S5 (supporting information) between the metalloporphyrin organic layer and Mg electrode is to improve the performance of the fabricated device significantly via lowering the electrode work function by the formed interfacial dipoles, which result in Fermi level pinning and band bending and suppressing the surface recombination of the photogenerated e–h pairs by manipulation of surface roughness defects. Furthermore, the LiF impact is declared in not only adding a tunneling barrier for the minor charge carriers which are responsible for the photodetector’s current but also limiting the metal electrode to diffuse into the organic layer during deposition, which would act as recombination centers and consequently, decrease the charge carrier’s mobility [97, 98].

On the other hand, it would be unreasonable not to evaluate the photodetection figures of merit of the manufactured heterogeneous photodetectors such as responsivity, R, and specific detectivity, D*, to be compared to the other organic/inorganic heterojunction photodetectors respecting the difference in architectural designs and working environments as displayed in Table 5. The responsivity measures the photodetection system’s input–output gain and equals the ratio between generated photocurrent density and the incident light intensity [90, 92]. The reverse bias dependence of responsivity of the designed metalloporphyrins photodetectors as a function of illumination intensity is revealed in Fig. 13a–d. As one can observe, the R-values descend gradually with decreasing bias voltage and significantly enhance with increasing the illuminating intensity for all the fabricated devices. The responsivity value of CuTPP/n-Si heterojunction is higher than that of other organic/n-Si heterojunctions such as Cu(acac)2/n-Si [92], α-6 T/n-Si [99], pentacene/n-Si [100], MG/n-Si [101], and TPPFT/n-Si [102]. But unfortunately, it possesses a lower responsivity than that of the TPD/n-Si heterojunction [90], NiTPP-C60/SiO2/Si [19], and graphene/n-Si [103].

Table 5 A comparison between Mg/LiF/MTPP/n-Si/Al and other organic/n-Si heterojunction photodetectors
Full size table
Figure 13
figure 13

Voltage dependence of responsivity of a CuTPP/n-Si, b NiTPP/n-Si, c FeTPPCl/n-Si, and d MnTPPCl/n-heterojunction at different illumination intensities

Full size image

Moreover, the capability of the fabricated devices to pick up the faintest light signals is illustrated from the obtained values of specific detectivity, D*, that are shown in Fig. 14a–d. The D*-values are calculated assuming that the dominant noise current contribution is shot noise and can be expressed as D* = R/(2qJd)1/2, where q is the electronic charge [90]. This shot noise results from the randomness of the generation and recombination process of the thermally generated e–h pairs [90, 92]. The calculated values of noise current, inoise, of the fabricated photodetectors are illustrated in Fig. S6 (supporting information) and are found to be in the range (0.1–0.5) pA with no observable impact of the light intensity variation, which assures the dominance of shot noise current mechanism and excluding Flicker and Johnson’s noise currents. Furthermore, noise equivalent power, NEP, that is given by NEP = (AB)1/2/D*, where A is the device area and B is detection bandwidth [90] and represents the minimum value of the light signal that can be detected to generate inoise is calculated for all fabricated photodetectors as shown in Fig. S7 (supporting information), where CuTPP/n-Si achieves the lowest value of NEP in the range of (0.025–0.25) nW/Hz1/2.

Figure 14
figure 14

Voltage dependence of specific detectivity of a CuTPP/n-Si, b NiTPP/n-Si, c FeTPPCl/n-Si, and d MnTPPCl/n-Si heterojunction at different illumination intensities

Full size image

Next, to guarantee the monolithic integrated performance of the fabricated photodetectors with the other electronic components of the photodetection system, the photodetector operation's linearity range is measured by the linear dynamic range parameter LDR, which should be determined [90, 92, 100]. Additionally, the ability of the fabricated photodetectors to differentiate between the noise and optical signals is measured in terms of signal-to-noise ratio, SNR. Figure 15a, b illustrates that CuTPP/n-Si heterojunction achieves the highest responsivity and specific detectivity to the probed light signal compared to other heterojunctions. The priority of CuTPP/n-Si heterojunction may be ascribed to the high absorption coefficient and low energy gap of CuTPP compared to other metalloporphyrins in this study, in addition to the low excitonic binding energy of CuTPP and low series resistance. Whereas CuTPP is the most common compound among the compounds under study conserving its molecular structure planarity, which may lead to an effective intermolecular π/π stacking resulting in a more facile delocalization of the photogenerated excitons [81]. Interestingly, the descended detectives of NiTPP/n-Si and MnTPPCl/n-Si are ascribed to the role of the high interface trap states that suppresses the dark current and improve the detectivity [96]. Nonetheless, FeTPPCl/n-Si heterojunction records the highest SNR and LDR as disclosed in Fig. 15c, d, which may be due to the low dark current in comparison to other heterojunctions.

Figure 15
figure 15

Responsivity, specific detectivity, signal to noise ratio, and linear dynamic range of the fabricated photodetectors at bias − 2.5 V and illumination 100 mW/cm.2

Full size image

It is well understood that as the photoresponse speed increases, the efficiency and thoroughness of the photodetector increases. Consequently, the response speed of the fabricated photodetectors is evaluated using a synchronized mechanically choppered light signal of frequency 0.05 Hz with variable intensities. Figure 16a–d depicts the transient temporal response signal of all the manufactured photodetectors with a repeatable and fastened response. The sensing signals show a stable reversible behavior in all sensing systems. The shape of the ON/OFF switchable behavior of photocurrent density illustrates that skin depth is higher than the width of the depletion region and that all devices have a low time constant, RC [104]. It worthy of comment is that CuTPP/n-Si achieves the strongest response signal (On/OFF ratio) upon illuminating the device, as shown in Fig. 17a, which increases its superiority. All the fabricated devices show a linear proportionality between incident light signal intensity and obtained ON/OFF ratio.

Figure 16
figure 16

Transient switching behavior of a CuTPP/n-Si, b NiTPP/n-Si, c FeTPPCl/n-Si, and d MnTPPCl/n-heterojunction at different illumination intensities

Full size image
Figure 17
figure 17

Illumination intensity dependence of a ON–OFF ratio, b rise time, c fall time of CuTPP/n-Si, NiTPP/n-Si, FeTPPCl/n-Si, and MnTPPCl/n-heterojunction and d rise and fall times at 100 mW/cm2 for the four compounds

Full size image

Meanwhile, the rise time, trise, and fall time, tfall, of the engineered photodetectors are calculated and noticed to be decreased as the illumination intensity decreases, as observed in Fig. 17b, c. The obvious effect of both rise and fall times on the change in the intensity of illumination incidents on the heterojunction can be attributed to improving the photoconductivity of the metalloporphyrin samples under study [90]. By comparison, CuTPP/n-Si records the lowest rise and fall times relative to other metalloporphyrins heterojunctions, as illustrated in Fig. 17d. The obtained rise and fall time values are lower than TPPFT/n-Si [101] and NiTPP-C60/SiO2/Si [102].

Conclusions

In this work, metalloporphyrin (CuTPP, NiTPP, FeTPPCl, and MnTPPCl) thin films were successfully deposited under the same conditions. The fabricated thin films exhibited an amorphous structure with a rough surface profile. Furthermore, the highly resolved XPS elemental analysis confirmed the stability of the deposited films and ensured the 50% loss of Cl in FeTPPCl and MnTPPCl films. The UV–Vis–NIR absorption spectra of the deposited films were fine analyzed and interpreted in terms of planarity and/or nonplanarity issues. The implemented heterojunction of copper tetraphenylporphyrin thin films with the n-type silicon showed a superior photodetection performance attributed to several reasons, the most important of which is its planar molecular structure's stability which resulted in lower values for the energy gap and lowered exciton binding energy. Also, considering the energy difference between the molecular orbitals energy of metalloporphyrins and silicon's conduction and valance bands energy, copper tetraphenylporphyrin showed the lowest potential barrier for the charges. Furthermore, it resulted in a higher response and speed to the light intensity change. Collectively, and by comparing the prepared photodetectors' merit with those reported in previous studies, the Mg/LiF/CuTPP/n-Si/Al photodetector showed a high and stable performance which makes this architecture a promising photodetector device of great importance in the field of optoelectronics. In our future work, we aim to improve the performance of the designed photodetectors, exploiting the effectiveness of plasmonic nanoparticles to improve photodetectors' response speed and efficiency. Moreover, we aim to investigate and optimize the performance of the metalloporphyrins for photodetection but in an organic/organic heterojunction architecture.