Porphyrin Nanotubes Overview

• Porphyrin nanotubes are diverse nanoscale architectures that couple porphyrinic functionality with cylindrical frameworks, including noncovalent CNT–TPP hybrids, template-synthesized porphyrin tubes, and embedded defects in CNTs.
• In CNT–TPP hybrids, free-base tetraphenylporphyrin molecules physisorb via π–π stacking onto carbon nanotubes, yielding binding energies up to 1.8 eV and effective energy-transfer properties.
• Template-directed porphyrin nanotubes and nitrogen-doped CNT defects exhibit tunable electronic and spin characteristics, such as oscillatory bandgaps and giant magnetoresistance, opening paths for optoelectronic and spintronic applications.

Searching arXiv for the specified papers and closely related work on porphyrin nanotubes. Searching arXiv for "porphyrin nanotubes tetraphenylporphyrin carbon nanotubes (Orellana, 2015, Orellana et al., 2015, 1803.01995, 0910.5619, Almeida et al., 2011)". Porphyrin nanotubes denote several distinct nanoscale architectures in which porphyrinic electronic structure is coupled to cylindrical or quasi-cylindrical frameworks. In one widely used sense, they are noncovalent hybrids in which free-base tetraphenylporphyrin (TPP) molecules physisorb on single-wall carbon nanotube (SWCNT) sidewalls, or fully cover them, through face-on $\pi$–$\pi$ stacking; related calculations also compare the same motif on graphene (Orellana et al., 2015, Orellana, 2015, 0910.5619). In a second sense, the term designates template-directed, bottom-up synthesized, fully $\pi$-conjugated molecular nanotubes built from fused porphyrin nanorings and characterized by unusual size-dependent bandgap oscillations (1803.01995). A third usage appears in nitrogen-doped carbon nanotubes containing embedded porphyrin-like $4N$ coordination pockets metalated with transition metals such as Fe, producing Heme B-like centers with spin-selective transport signatures (Almeida et al., 2011). The shared terminology therefore reflects porphyrinic functionality and nanotubular geometry, but not a single structural class.

1. Terminology and structural scope

The phrase “porphyrin nanotubes” is not univocal in the literature represented here. It covers at least three structurally distinct systems: noncovalent CNT–TPP adsorbates, template-directed $\pi$-conjugated porphyrin-only nanotubes, and porphyrin-like coordination defects embedded in CNT walls. A common source of confusion is the assumption that all porphyrin nanotubes are self-assembled tubes made only from porphyrins. In the CNT–TPP literature, this is explicitly not the case: the nanotube is carbon, and the porphyrin forms a noncovalent coating rather than the tube itself (Orellana, 2015, Almeida et al., 2011).

Class	Defining structural feature	Representative source
$\pi$-stacked CNT–TPP hybrids	Free-base TPP physisorbed on SWCNT sidewalls, from isolated adsorbates to full monolayer wraps	(Orellana et al., 2015, Orellana, 2015, 0910.5619)
Template-directed $\pi$-conjugated PNTs	Fused porphyrin nanorings assembled into fully conjugated molecular nanotubes	(1803.01995)
Embedded porphyrin-like CNT defects	Transition-metal centers coordinated by a $4N$ divacancy in an N-doped CNT wall	(Almeida et al., 2011)

For CNT-based hybrids, the relevant nanotube geometry is indexed by chirality $(n,m)$, with diameter

$d = \frac{a}{\pi}\sqrt{n^2 + nm + m^2},$

where $\pi$0 Å. In the broad SWCNT–TPP survey, 42 species with $\pi$1 and $\pi$2 were studied, spanning metallic and semiconducting tubes over a diameter range of 3.9–16.6 Å (Orellana et al., 2015). In the template-directed molecular series, “size” instead denotes the number of porphyrin subunits around the circumference, $\pi$3, with modeled outer diameters spanning approximately 0.6 to 15.4 nm (1803.01995). The embedded-defect usage is again different: the nanotube is an armchair $\pi$4 CN$\pi$5 tube, and the porphyrin-like character resides in a square/pyridinic $\pi$6 coordination pocket in the wall (Almeida et al., 2011).

2. Noncovalent CNT–TPP architectures and assembly

In the noncovalent class, free-base TPP adsorbs on graphitic carbon by dispersive $\pi$7–$\pi$8 interactions without covalent disruption of the $\pi$9 network. Spin-unpolarized DFT calculations using SIESTA, a DZP basis, norm-conserving pseudopotentials, and nonlocal vdW-DF were used to analyze isolated TPP adsorption on 42 SWCNT species and on graphene (Orellana et al., 2015). The adsorption energy was defined with BSSE correction as

$\pi$0

with exothermic binding corresponding to $\pi$1; the paper reports positive magnitudes $\pi$2 (Orellana et al., 2015).

Three low-energy TPP registries were examined on CNTs: N–H bonds perpendicular to the tube axis, parallel to the tube axis, and at $\pi$3. The lowest-energy structure has the N–H bonds perpendicular to the tube axis, but the energy differences among the three are $\pi$4 eV, so multiple orientations are thermally accessible. The adsorption distance, defined between the CNT surface and the TPP nitrogen atoms, lies in 3.0–3.3 Å for CNTs; representative values include 3.000 Å for $\pi$5, 3.114 Å for $\pi$6, and 3.180 Å for $\pi$7 (Orellana et al., 2015).

Across the 42 isolated-adsorbate CNT systems, $\pi$8 spans 1.1–1.8 eV and increases quasi-linearly with tube diameter. Representative values are 1.11 eV for $\pi$9, 1.39 eV for $4N$0, 1.54 eV for $4N$1, 1.59 eV for $4N$2, and 1.82 eV for $4N$3. When plotted versus chiral angle, any apparent correlation disappears once diameter is controlled. The dominant control parameter is therefore contact area rather than chirality, with local deviations arising from registry and steric repulsion between phenyl-ring hydrogens and underlying carbon atoms (Orellana et al., 2015).

A higher-coverage regime was analyzed separately for single-wall and double-wall CNTs fully covered by TPP. In those geometries, TPPs form compact rings around the tube, with interdigitated phenyl groups, and full coverage is thermodynamically favored because neighboring porphyrins add roughly 0.2 eV/TPP in stabilization through additional vdW attractions. Binding energies per TPP reach 1.95 eV for $4N$4-8TPP, 2.10 eV for $4N$5-5TPP, 2.26 eV for $4N$6-5TPP, 2.20 eV for $4N$7-5TPP, and 2.24 eV for $4N$8-5TPP, with equilibrium adsorption distances near 3.1–3.2 Å at 0 K (Orellana, 2015).

Experimental assembly of $4N$9-stacked SWCNT/TPP complexes was demonstrated by swelling sodium cholate micelles with dichloromethane (DCM). Isolated CoMoCAT SWCNTs were dispersed in pH 8 Normadose buffer with 2 wt% sodium cholate, and TPP dissolved in DCM was introduced into the micellar suspension, followed by 2 h tip sonication at 12°C. The DCM/water ratio controlled functionalization: at approximately 2% no functionalization was detected, at approximately 7% the Soret response was weak, and at $\pi$0–27% robust functionalization and aggregate formation in micelles were observed; 2.5 mL SWCNT suspension plus 0.85 mL DCM, corresponding to approximately 34% v/v, was used for optimal synthesis (0910.5619).

The micelle-swelling mechanism is explicit. DCM swells the hydrophobic core of the sodium cholate micelle, carries non-water-soluble TPP into direct contact with the nanotube sidewall, then evaporates post-sonication, leaving stable porphyrin–SWCNT complexes trapped in the micelle core. Saturation of the 438 nm stacked-TPP band above approximately 0.14 $\pi$1mol TPP under the stated conditions indicates a finite number of accessible adsorption sites per tube and micellar confinement, consistent with a Langmuir-like adsorption process (0910.5619).

3. Electronic structure, spectroscopy, and energy transfer in physisorbed hybrids

For isolated TPP on SWCNTs, the principal electronic outcome is weak perturbation of the nanotube host together with preservation of the porphyrin’s main optical absorptions. In the broad 42-tube DFT survey, no significant charge transfer is reported for CNTs; the tube band structure and gap, if semiconducting, are largely preserved. By contrast, graphene–TPP shows a qualitative doping signature: the Dirac point lies below the Fermi level after adsorption, indicating electron transfer from TPP to graphene and an $\pi$2-type doping mechanism without gap opening or disruption of the graphene band topology (Orellana et al., 2015).

The optical response was computed within first-order time-dependent perturbation theory from the imaginary part of the dielectric function,

$\pi$3

with a Gaussian broadening of 0.06 eV. This independent-particle DFT treatment omits quasiparticle and excitonic effects, so absolute peak positions are redshifted relative to experiment by approximately 0.6–0.8 eV, but relative trends and band separations are meaningful (Orellana et al., 2015, Orellana, 2015).

For the $\pi$4-TPP reference system, the computed TPP bands occur at 1.60 eV (Q) and 2.04 eV (B/Soret), separated by 0.44 eV, close to the experimental Q–B separation of 0.474 eV although redshifted relative to the measured 2.357 eV and 2.831 eV values. Across semiconducting, metallic, and semimetallic CNTs, the TPP Q and B bands remain at essentially the same energies as in the isolated molecule within the same computational setup. This invariance is one of the central results of the physisorbed-hybrid literature: adsorption on CNTs does not shift TPP’s main absorptions in any systematic way (Orellana et al., 2015).

At full coverage, the spectra of semiconducting CNT–TPP compounds are dominated by strong TPP-derived features near 1.5 eV ($\pi$5) and 2.0 eV ($\pi$6), and their intensities scale with TPP concentration. Specific shifts depend on tube and coverage. For $\pi$7-8TPP, the $\pi$8 peak is redshifted by approximately 0.1 eV relative to free TPP, whereas in $\pi$9-5TPP and $\pi$0-5TPP the $\pi$1 band is redshifted and the Soret band remains nearly unchanged. Single-wall and double-wall systems show closely similar TPP-dominated optical signatures, with only small screening-induced shifts in double-wall tubes (Orellana, 2015).

Graphene provides an upper bound in noncovalent binding and a distinct optical perturbation. For graphene–TPP, $\pi$2 eV at $\pi$3 Å, approximately twice the strongest CNT value, and the TPP-derived absorptions appear at approximately 1.57 eV and 1.96 eV, corresponding to a redshift of approximately 0.05 eV relative to isolated TPP under the same conditions. The redshift is attributed to stronger $\pi$4–$\pi$5 coupling and phenyl-ring distortion on graphene (Orellana et al., 2015).

Experimentally, SWCNT/TPP micelle complexes show a dual Soret response: 438 nm for stacked TPP and 420 nm for free TPP in micelles. The nanotube $\pi$6 lines redshift by approximately 7 nm in absorption and the PL peak positions redshift by approximately 20 nm, both attributed to dielectric screening. In photoluminescence excitation, SWCNT/TPP complexes detected at 1007 nm show a strong additional peak at 441 nm coincident with the stacked-TPP Soret absorption, while the SWCNT $\pi$7 line appears near 580 nm and is redshifted by approximately 10 nm relative to the reference. The comparable amplitudes of the PLE peaks at 441 nm and at the SWCNT $\pi$8 were taken as evidence of efficient energy transfer from TPP to SWCNT (0910.5619).

The energy-transfer paper treats the process as energy transfer without definitively distinguishing Förster, Dexter, or charge-transfer pathways. Its stated interpretation is that the preserved SWCNT luminescence and dielectric-driven spectral shifts support an excitonic FRET-like mechanism rather than net charge transfer, while donor lifetimes and fluorescence quenching were not measured (0910.5619). This suggests that, in physisorbed CNT–TPP assemblies, optical sensitization can be substantial even when ground-state electronic restructuring is limited.

4. Template-directed $\pi$9-conjugated porphyrin nanotubes

A different porphyrin nanotube class is obtained by template-directed, bottom-up synthesis from fused porphyrin nanorings. In the work analyzed by large-scale DFT, the underlying synthetic context is the Vernier template-directed methodology associated with the Anderson group, which enables precise control of ring size up to diameters of approximately 10 nm and subsequent assembly into fully conjugated, three-dimensional molecular nanotubes. Computational models linked 2 to 36 porphyrin subunits with acetylenic linkages into one-dimensionally periodic tubes, denoted $\pi$0 according to the number of porphyrin subunits around the circumference (1803.01995).

The defining electronic feature of these PNTs is an unusual oscillation of the bandgap with size. The gap is defined as

$\pi$1

Contrary to conventional quantum-confinement expectations, even-numbered PNTs ($\pi$2, $\pi$3, $\pi$4) show bandgaps that increase with size, whereas odd-numbered PNTs ($\pi$5, $\pi$6, $\pi$7) show bandgaps that decrease with size. The global variation across the series reaches 0.32 eV and asymptotically approaches 0.81 eV, equal to the bandgap of the parent planar porphyrin sheet (1803.01995).

The mechanism is a size-dependent alternation of aromatic and non-aromatic circumferential conjugation. Each porphyrin contributes 14 $\pi$8-electrons to the continuous ring pathway, so a tube with $\pi$9 porphyrins contains $4N$0 $4N$1-electrons around the circumference. For odd $4N$2, $4N$3, satisfying the Hückel aromaticity class $4N$4; for even $4N$5, $4N$6, yielding the non-aromatic class $4N$7. The more negative valence band maxima of odd-numbered tubes track this aromatic stabilization, and the oscillation of the band edges generates the oscillatory $4N$8 behavior (1803.01995).

Band-structure calculations show that all such PNTs possess a direct bandgap at the $4N$9 point. The conduction band near the gap is nearly dispersionless, implying extremely localized electronic states and, correspondingly, a very large electron effective mass. Orbital-symmetry analyses for $(n,m)$0 and $(n,m)$1 show different symmetries for HOMO and LUMO at $(n,m)$2, implying optically allowed, “bright” transitions for all tubes studied (1803.01995).

Polarization selection follows the parity-dependent spatial character of the LUMO. For odd-numbered tubes such as $(n,m)$3, the LUMO is localized on staves parallel to the tube axis, and excitation occurs for light polarized along the axis. For even-numbered tubes such as $(n,m)$4, the LUMO localizes around the circumference, and excitation occurs for light polarized perpendicular to the axis, including circular polarization (1803.01995). This polarization anisotropy is not described as a minor correction, but as a structural consequence of the orbital topology.

The same oscillatory band-edge landscape permits heterojunction engineering within a single chemical family. Consecutive odd-numbered pairs such as $(n,m)$5 show type I alignment, whereas consecutive even-numbered pairs such as $(n,m)$6 show type II alignment. The reported conduction-band offset for $(n,m)$7 is $(n,m)$8 eV with PBE, and a larger-size contrast such as $(n,m)$9 yields $d = \frac{a}{\pi}\sqrt{n^2 + nm + m^2},$0 eV with HSE06/DZP (1803.01995). The paper emphasizes that both type I and type II donor–acceptor $d = \frac{a}{\pi}\sqrt{n^2 + nm + m^2},$1–$d = \frac{a}{\pi}\sqrt{n^2 + nm + m^2},$2 heterojunctions are possible in these template-directed porphyrin nanotubes, unlike the conventional CNT case described there.

5. Embedded porphyrin-like centers in nitrogen-doped carbon nanotubes

A third, chemically distinct meaning of porphyrin nanotube arises in nitrogen-doped CNTs containing porphyrin-like defects. Here the structural motif is the “porphyrin-like” four-nitrogen divacancy, in which two adjacent carbon vacancies are surrounded by four substitutional nitrogen atoms to form a square/pyridinic $d = \frac{a}{\pi}\sqrt{n^2 + nm + m^2},$3 pocket analogous to a porphyrin macrocycle. Transition-metal atoms placed in this site bind strongly in a porphyrin-like fashion. For Fe, the resulting Fe–$d = \frac{a}{\pi}\sqrt{n^2 + nm + m^2},$4 center is explicitly compared with Heme B and carries a localized net magnetic moment of $d = \frac{a}{\pi}\sqrt{n^2 + nm + m^2},$5 (Almeida et al., 2011).

The reported binding energies relative to isolated transition-metal atoms are strongly negative: Fe $d = \frac{a}{\pi}\sqrt{n^2 + nm + m^2},$6 eV, Co $d = \frac{a}{\pi}\sqrt{n^2 + nm + m^2},$7 eV, Mn $d = \frac{a}{\pi}\sqrt{n^2 + nm + m^2},$8 eV, and Ni $d = \frac{a}{\pi}\sqrt{n^2 + nm + m^2},$9 eV. Fe–$\pi$00 is both the most stable center and the one with the highest low-temperature single-defect polarization, 8.95%, compared with 1.46% for Co–$\pi$01, 2.70% for Mn–$\pi$02, and 0.00% for Ni–$\pi$03 (Almeida et al., 2011). The mechanism is exchange-split, localized $\pi$04-like impurity states hybridized with CNT bands, with minority-spin Fe states lying closer to $\pi$05 than majority-spin states.

Transport was modeled for metallic armchair $\pi$06 CN$\pi$07 nanotubes using DFT within SIESTA and recursive Green’s functions. In long disordered tubes with randomly positioned Fe–$\pi$08 centers at 0.65% defect concentration and 3 K, the spin-dependent conductance decays exponentially with length,

$\pi$09

For fully aligned moments, the extracted localization lengths are $\pi$10 Å and $\pi$11 Å; for 80% alignment, $\pi$12 Å and $\pi$13 Å; for 50% alignment, $\pi$14 Å and $\pi$15 Å (Almeida et al., 2011).

Because the minority-spin channel localizes more strongly when moments are aligned, the conductance polarization increases sharply with length. The reported result is that polarization approaches approximately 100% beyond approximately 200 nm for 100% alignment and remains nearly 100% by approximately 700 nm even for 80% alignment, all with nonmagnetic electrodes (Almeida et al., 2011). The disorder-induced magnetoresistance,

$\pi$16

reaches up to approximately 20000% for fully aligned moments and approximately 1000% for 80% alignment (Almeida et al., 2011).

This embedded-defect usage should be distinguished from both template-directed porphyrin-only nanotubes and noncovalently coated CNT–TPP hybrids. The porphyrin-like site is covalently integrated into the CNT wall and acts as a reactive coordination center rather than an adsorbed chromophore. The paper’s proposed device consequences are therefore spin-filtering and magnetoresistive behavior rather than the optical sensitization emphasized in TPP-coated CNTs (Almeida et al., 2011).

6. Design principles, limitations, and recurring issues

Several design rules recur across the noncovalent CNT–TPP literature. First, diameter is the main energetic control variable for isolated TPP adsorption on SWCNTs; chirality does not provide an independent binding selector once diameter is fixed (Orellana et al., 2015). Second, full coverage raises the binding energy per TPP through lateral phenyl–phenyl vdW interactions and yields compact interlocked porphyrin rings around the nanotube (Orellana, 2015). Third, preserving the porphyrin’s Q and B bands while tuning CNT excitonic transitions into resonance is treated as a practical route to light harvesting, photodetection, or emission control, with specific SWCNTs having $\pi$17 transitions near TPP B or Q bands (Orellana et al., 2015). Fourth, graphene offers stronger adsorption and qualitative $\pi$18-type doping while retaining the graphene lattice and band topology (Orellana et al., 2015).

The experimental micelle-swelling route adds process-level constraints. Functionalization depends on the DCM/water ratio, sonication time and energy, temperature control at 12°C to limit DCM evaporation during swelling, and the quality of the starting dispersion of isolated SWCNTs in sodium cholate micelles. Above the empirical saturation threshold of approximately 0.14 $\pi$19mol TPP under the reported conditions, the 420 nm band continues to grow because excess TPP aggregates in micelles (0910.5619). This suggests that coverage optimization is inseparable from micellar transport and aggregation control.

The template-directed molecular PNTs follow a different design logic. There, the primary tuning parameter is circumference parity rather than chirality or adsorption registry. Odd and even $\pi$20 tubes alternate between aromatic and non-aromatic circumferential conjugation, which determines whether $\pi$21 decreases or increases with size and whether a given pair forms a type I or type II heterojunction (1803.01995). The practical implication stated in that work is that monodisperse, Vernier-templated synthesis can be used to engineer donor–acceptor band offsets within a single material family.

The embedded Fe–$\pi$22 systems impose yet another constraint set. Strong binding and high spin selectivity favor Fe, but operation is tied to magnetic alignment. The paper’s ideal paramagnet estimate states that room-temperature fields would be impractically large, approximately 200 T, whereas at 3 K roughly 2 T yields approximately 80% alignment and roughly 5 T yields approximately 95% alignment (Almeida et al., 2011). A plausible implication is that the reported spin-filtering and giant magnetoresistance are most directly relevant to cryogenic devices.

Methodological limits are explicit throughout the cited literature. Independent-particle DFT omits GW/BSE quasiparticle and excitonic effects in CNT/TPP and graphene/TPP optical spectra, so absolute peak positions are redshifted by approximately 0.6–0.8 eV and excitonic fine structure is absent (Orellana et al., 2015, Orellana, 2015). The micelle-swelling experiments did not measure donor lifetimes, fluorescence quenching, or quantitative Förster parameters, so transfer efficiency is assessed qualitatively rather than by extracted $\pi$23 or $\pi$24 (0910.5619). The template-directed PNT calculations use idealized one-dimensional periodic structures in vacuum, so screening by templates, substrates, or supramolecular stacking is not included (1803.01995). The embedded-defect transport analysis neglects spin–orbit coupling and phonon scattering and assumes coherent transport in the two-spin-fluid approximation (Almeida et al., 2011).

Taken together, these studies define porphyrin nanotubes not as a single material, but as a family of porphyrin-enabled nanotubular systems with three distinct operational regimes. Noncovalent CNT–TPP hybrids emphasize stable adsorption, preserved dye spectroscopy, and energy-transfer functionality; template-directed porphyrin nanotubes emphasize circumference-controlled aromaticity, direct bright gaps, and heterojunction design; embedded porphyrin-like CNT defects emphasize covalent Fe–$\pi$25 stabilization, spin-dependent localization, and magnetotransport (Orellana et al., 2015, Orellana, 2015, 0910.5619, 1803.01995, Almeida et al., 2011).