Direct synthesis of dilithium tetraphenylporphyrin: facile reaction of a free-base porphyrin with vapor-deposited lithium

Article abstract of DOI:10.1039/c9cc07170a

A solvent-free dilithium porphyrin was synthesized by direct reaction of free-base meso-tetraphenylporphyrin with elemental lithium in ultra-high vacuum. The reaction product dilithium tetraphenylporphyrin was studied by temperature-programmed desorption

Full text of DOI:10.1039/c9cc07170a

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Direct synthesis of dilithium tetraphenylporphyrin:
facile reaction of a free-base porphyrin with
vapor-deposited lithium†
Cite this: Chem. Commun., 2019,
55, 13665
Received 13th September 2019,
Accepted 15th October 2019
Maik Schoniger, ‡ Stefan R. Kachel,‡ Jan Herritsch,
Philipp Schroder,
¨
¨
Mark Hutter and J. Michael Gottfried
*
DOI: 10.1039/c9cc07170a
rsc.li/chemcomm
A solvent-free dilithium porphyrin was synthesized by direct reaction
Here, we report that solvent-free dilithium tetraphenylpor-
of free-base meso-tetraphenylporphyrin with elemental lithium in phyrin (Li2TPP) can be prepared by vapor deposition of lithium
ultra-high vacuum. The reaction product dilithium tetraphenylpor- onto free-base meso-tetraphenylporphyrin (H2TPP) in ultra-high
phyrin was studied by temperature-programmed desorption mass vacuum (UHV), by the reaction shown in Fig. 1. Unlike the
spectrometry (TPD-MS) and hard X-ray photoelectron spectroscopy very limited reaction depths of transition metals in bulk
(HAXPES). The solid-state reaction is thermodynamically favored, porphyrins,15,16 or the metalation of monolayers,4,17–19 we find
according to density functional theory (DFT) calculations.
that Li diffuses readily into the solid porphyrin film, which
makes the reaction useful for synthetic purposes. The bulk
Porphyrins, the ‘colors of life’,1 are widespread in nature and reaction between metallic Li and organic molecules is also
find numerous technological applications.2–4 While transition- relevant for the field of organic electronics, where metals with
metal porphyrins can be very stable with respect to metal low work functions are vapor-deposited onto organic semi-
exchange, alkali metal porphyrins (AMPs) with electron-rich conductors as electron-injecting electrodes and intermixing or
ligands such as meso-tetraphenylporphyrin (TPP) are considered reaction of the layers is usually not desirable.20–22 The reaction
as labile and water-sensitive complexes.5 Although AMPs were of Li with organic materials also plays a role in Li-based
observed as early as 1938,6 they were mainly characterized in electrochemical energy storage, in which the formation of the
solution or as solvates. The lability of AMPs in solution has been solid–electrolyte interphase (SEI) and the diffusion of lithium in
used for the synthesis of other metalloporphyrins by metal the SEI is crucial for device performance.23,24 Porphyrin-based
exchange.7–10 The first X-ray structure of an AMP solvate was electrodes have also proven to be efficient electrode materials in
obtained for the tetrahydrofuran (THF) adduct of dilithium lithium batteries.25
octaethylporphyrin, Li2OEP(THF)4, formed by reaction of
The Li2TPP layers obtained by the reaction shown in Fig. 1
octaethylporphyrin (H2OEP) with LiN(SiMe3)2.11 The solid consists were investigated by temperature-programmed desorption mass
of a [Li(THF)4]+ cation and a [Li(OEP)]ꢀ anion with a square-planar
coordination of Li in the center of the porphyrin ring.11 In other
AMP solvates such as Li2TPP(OEt2)2, both Li ions are bound to
opposite sides of the porphyrin ring in a square-bipyramidal
fashion and carry an additional diethyl ether (OEt2) ligand.12 This
bipyramidal structure has also been reported for dilithium
porphyrins in solution and for solid Na and K porphyrins.13
Besides these structurally well-characterized solvates, the two
solvent-free AMPs Li2TPP and Li2OEP have been obtained from
solution, but have not been studied in detail.14
Fig. 1 Direct metalation of meso-tetraphenylporphyrin (H2TPP) with two
equivalents of elemental Li, resulting in the formation of dilithium tetra-
phenylporphyrin (Li2TPP) and H2. The figure shows the geometry-optimized
gas-phase structures calculated by DFT. The distorted square-bipyramidal
polyhedron formed by the lithium and nitrogen atoms in Li2TPP is
highlighted.
Department of Chemistry, Philipps University Marburg, Hans-Meerwein-Straße 4,
35032 Marburg, Germany. E-mail: michael.gottfried@chemie.uni-marburg.de
† Electronic supplementary information (ESI) available: Experimental and com-
putational details, additional TPD-MS and XPS data, XPS information depths,
Born–Haber cycle, additional DFT results. See DOI: 10.1039/c9cc07170a
‡ These authors contributed equally to this study and share first authorship.
This journal is ©The Royal Society of Chemistry 2019
Chem. Commun., 2019, 55, 13665--13668 | 13665

View Article Online
Communication
ChemComm
spectrometry (TPD-MS) and hard X-ray photoelectron spectro- of 565 K, which exceeds that of H2TPP by 80 K, and the higher
scopy (HAXPES). Complementary density functional theory (DFT) desorption energy of 234 kJ molꢀ1 (Fig. 2d) reveal increased
calculations provide insight into the energetics of the reaction.
intermolecular interactions, compared to H2TPP. The H2TPP
Before the reaction between H2TPP and Li was studied, signal has completely disappeared, which shows that the free-
reference TPD-MS measurements were performed with pristine base porphyrin reacts completely. Weak additional features
H2TPP. For this purpose, a film of H2TPP with a thickness of appear at 619 and 637 amu. These masses correspond to LiTPP
20 nm was vapor-deposited onto a clean Au(111) surface. and LiTPPꢁH2O, respectively. The LiTPP is most likely formed
Subsequently, a TPD-MS experiment was performed by heating by fragmentation of Li2TPP in the mass spectrometer, because
the sample with a constant rate of 0.1 K sꢀ1 between 300 and the desorption traces of Li2TPP and LiTPP have the same
700 K while rapidly recording mass spectra, as described in the temperature dependence (see Fig. S2 in the ESI†) and their
ESI.† The resulting temperature-resolved mass spectra are intensities have a constant ratio of E12 : 1. The LiTPPꢁH2O
shown as TPD-MS maps in Fig. 2a and Fig. S1a in the ESI.† adduct (E5% of the Li2TPP signal) is attributed to reaction with
Around 485 K a dominant peak is observed in a mass range that residual water. The small Li2TPP-related signal around 500 K is
indicates desorption of intact H2TPP (614 amu). Arrhenius plot due to the formation of small crystallites with lower desorption
analysis of the zeroth-order desorption trace (Fig. 2e) yields a energy (see Arrhenius analysis in Fig. S3 in the ESI†), in agreement
desorption activation energy of 200 kJ molꢀ1, which is in the with previous work.27 It is worth mentioning that the LiTPPꢁH2O
range of literature values for the sublimation enthalpy (between adduct is only observed at temperatures where the crystallites
111 and 240 kJ molꢀ1).26
desorb.
To investigate whether even thicker H2TPP films can be fully
To study the reaction between H2TPP and Li, a fresh 20 nm
thick film of H2TPP was deposited onto the clean Au(111) converted, 10 nm Li were deposited onto a 50 nm H2TPP film.
surface, followed by the deposition of 1 nm Li at 300 K. This An over-stoichiometric amount of Li was chosen to ensure
thickness of Li is close (within 5%) to the stoichiometric complete reaction. The TPD-MS map in Fig. 2c shows that the
amount required for complete conversion of the H2TPP film main product is again Li2TPP, with the water adduct Li2TPPꢁH2O
to Li2TPP. After this preparation, a TPD-MS experiment was (644 amu) as a side product (E13%). Analysis of the individual TPD
performed. The results in Fig. 2b show a dominant signal traces (see Fig. S2b in the ESI†) indicates that the film substantially
around the mass of Li2TPP (626 amu), indicating the formation recrystallizes during the heating,27 which leads to the observed
of the dilithium complex in high yield. The peak temperature broadening of the Li2TPP peak. Again, no traces of residual H2TPP
were found, indicating complete reaction.
The mass spectra presented above show that free-base H2TPP
reacts completely with Li forming Li2TPP. However, it remains
unclear whether the reaction occurs already at room temperature
or is induced by the elevated temperatures during the TPD-MS
experiment. To clarify this question, additional experiments with
X-ray photoelectron spectroscopy (XPS) were performed.
The XPS measurements were carried out with synchrotron
radiation in the hard X-ray regime between 2 and 7 keV, to take
advantage of the increased information depth of photoelectrons
with high kinetic energies.16,28 In this way, the composition of
the film can be probed over the entire thickness and not only at
the surface, as is the case for conventional XPS. At the highest
photon energy (Eph) of 7 keV, the information depth of N 1s
photoelectrons is 33.6 nm for H2TPP and 32.7 nm for Li2TPP, as
was calculated by the Gries approach.16,29 The details of the
calculation and the information depths for the other photon
energies are given in the ESI.† The XPS measurements were
performed with a H2TPP film of 24 nm thickness, which is
substantially smaller than the maximum information depth. By
variation of Eph (and thus, the information depth), the depth
homogeneity of the reacted film can additionally be examined.
Fig. 2 TPD-MS maps for (a) 20 nm H2TPP, (b) 1 nm Li on 20 nm H2TPP,
This film was prepared on a Si(001) surface, as is described in
and (c) 10 nm Li on 50 nm H2TPP, each on Au(111), in the relevant mass
region from 600 to 650 amu. Data for other mass ranges are shown in the detail in the ESI.†
ESI.† The spectra were normalized to the maximum intensity of the H2TPP
The resulting N 1s XP spectra are shown in Fig. 3. The
signal (a). The heating rate was 0.1 K sꢀ1 and one mass range cycle took
8 seconds (see the ESI† for further experimental details). Arrhenius plot
analysis for (d) Li2TPP and (e) H2TPP, with the desorption rates rdes divided
by the mass spectrometer unit (counts per second, for further details see
spectrum of the pristine H2TPP film (Fig. 3a) consists of two peaks,
which are attributed to the pyrrolic (–N(H)–, 400.2 eV) and iminic
(–NQ, 398.1 eV) nitrogen species,4,30 see Fig. 1. After deposition of
the ESI†), yields desorption energies of 234 and 200 kJ molꢀ1, respectively. Li at room temperature (Fig. 3b–e), a new peak appears at a
13666 | Chem. Commun., 2019, 55, 13665--13668
This journal is ©The Royal Society of Chemistry 2019

View Article Online
ChemComm
Communication
For the reaction shown in Fig. 1, a total energy of DEDFT
=
ꢀ499.8 kJ molꢀ1 was obtained, indicating that the gas-phase
process is strongly exothermic. In the following, we use this
gas-phase value to estimate the energy for the corresponding
solid-state reaction by means of a Born–Haber (BH) cycle (see
Fig. S6 in the ESI†), which includes the sublimation enthalpy
of Li metal and the desorption energies of H2TPP and Li2TPP
as obtained from TPD-MS above. The BH cycle yields an
approximate reaction energy of ꢀ215 kJ molꢀ1, which is still
strongly exothermic. Another driving force of the solid-state
reaction is the entropy gain from the release of gaseous H2
ꢀ
ꢁ
Sgꢂas;1 bar ¼ 130:68 J molꢀ1 Kꢀ1 at 298:15 K .31 Energetic and
entropic contributions add up to a standard Gibbs energy of
approximately ꢀ254 kJ molꢀ1, indicating that the formation of
Li2TPP is thermodynamically favored even in the solid state.
Vapor deposition should therefore not be necessary to induce
the reaction, except to overcome kinetic barriers. The calculated
minimum energy structure of the Li2TPP complex is shown in
Fig. S7 in the ESI.† The Li ions are located on both sides of the
porphyrin ring, but not above (or below) the center of the
porphyrin ring, in agreement with previous theory work.32
The frontier orbitals HOMOꢀ1, HOMO, LUMO and LUMO+1
are exclusively localized at the porphyrin ring; the Li ions do not
contribute (see Fig. S8 in the ESI†).
Fig. 3 N 1s X-ray photoelectron spectra of (a) H2TPP and (b)–(e) Li2TPP
formed by deposition of Li onto a 24 nm thick H2TPP layer, taken with the
indicated photon energies.
binding energy (BE) of 399.6 eV, while the two H2TPP-related peaks
disappear. The presence of a single N 1s peak indicates that all
four nitrogen atoms are chemically equivalent, which is the case in
the metal complex.4,18,19 The information depth of the experiment
exceeds the film thickness also after deposition of Li, as can be
seen from the Si 1s signal related to the Si(001) substrate in Fig. S4
in the ESI.† The formation of Li2TPP is also supported by the
related Li 1s XPS data in Fig. S5 in the ESI,† which show a
chemical shift to higher BE in line with the oxidation of metallic
Li to Li+ ions.
The absence of any H2TPP-related signals at a photon energy
of 7 keV, where the information depth exceeds the film thickness,
proves that the H2TPP film reacts over its entire thickness at
room temperature. For the photon energies of 2, 2.5 and 3 keV
(Fig. 3b–d), the information depths in Li2TPP are 10.1, 12.6 and
15.0 nm (see Table S3 in the ESI†), respectively, i.e., they are
smaller than the film thickness. Still, position and general shape
of the N 1s peaks are the same as for the 7 keV spectrum,
indicating that the reaction is homogeneous over the entire depth
of the film.
The fact that the H2TPP film reacts over its entire thickness
of 24 nm (or even 50 nm in the TPD-MS experiment) is exceptional
when compared to related reactions with transition metals. For
example, when Fe or Co are deposited onto H2TPP films at room
temperature, formation of the related FeTPP and CoTPP complexes
is limited to a depth of approximately 1 nm.16 Additionally
deposited metal forms a layer of metallic Fe or Co on-top of the
organic film. It can be concluded that the room-temperature
mobility of Li atoms in the solid porphyrin is much higher than
that of typical transition metals.
In conclusion, we have demonstrated that dilithium tetra-
phenylporphyrin (Li2TPP) can be prepared in solvent-free form
by direct reaction of meso-tetraphenylporphyrin (H2TPP) with
vapor-deposited metallic Li at room temperature. DFT calculations
show that the reaction is strongly exothermic in the gas phase. A
negative standard Gibbs energy for the solid-state reaction con-
firms that the reaction is thermodynamically favored also in the
condensed phase. Our results may pave the way to simple solid-
state syntheses of alkali metal porphyrins and thus increase the
availability of these compounds as reactants for organic and
inorganic synthesis. The exceptionally high mobility of Li in
the organic material is relevant for the deposition of metal
electrodes in organic electronic devices and the transport of Li
in electrochemical energy storage.
Financial support by the Deutsche Forschungsgemeinschaft
(DFG, German Research Foundation) through project number
223848855 – SFB 1083 is gratefully acknowledged. We thank the
Helmholtz-Zentrum Berlin for allocation of synchrotron radiation
beamtime at BESSY II and financial support. We thank M. Albus,
C. Zu¨lch and F. Fillsack for assistance during the measurements.
We thank the HRZ Marburg for computational resources.
Conflicts of interest
There are no conflicts to declare.
Notes and references
1 L. R. Milgrom, The Colors of Life, Oxford University Press, 1997.
2 Handbook of Porphyrin Science, ed. K. M. Kadish, K. M. Smith and
R. Guilard, Singapore, 2010.
To evaluate whether the direct formation of Li2TPP is thermo-
dynamically feasible, gas-phase DFT calculations were performed.
¨
3 W. Auwarter, D. Ecija, F. Klappenberger and J. V. Barth, Nat. Chem.,
2015, 7, 105.
This journal is ©The Royal Society of Chemistry 2019
Chem. Commun., 2019, 55, 13665--13668 | 13667

View Article Online
Communication
ChemComm
4 J. M. Gottfried, Surf. Sci. Rep., 2015, 70, 259.
19 H. Marbach, Acc. Chem. Res., 2015, 48, 2649.
5 J. Arnold, in The Porphyrin Handbook, ed. K. M. Kadish, K. M. Smith 20 F. Bebensee, M. Schmid, H. P. Steinru¨ck, C. T. Campbell and
and R. Guilard, Inorganic, Organometallic and Coordination
Chemistry, Academic Press, 2000, vol. 3, ch. 17, p. 113.
6 H. Fischer and K. Herrle, Hoppe-Seyler’s Z. Physiol. Chem., 1938,
251, 85.
J. M. Gottfried, J. Am. Chem. Soc., 2010, 132, 12163.
21 F. Bebensee, J. F. Zhu, J. H. Baricuatro, J. A. Farmer, Y. Bai, H. P.
Steinru¨ck, C. T. Campbell and J. M. Gottfried, Langmuir, 2010,
26, 9632.
22 J. Zhu, F. Bebensee, W. Hieringer, W. Zhao, J. H. Baricuatro, J. A.
Farmer, Y. Bai, H.-P. Steinru¨ck, J. M. Gottfried and C. T. Campbell,
J. Am. Chem. Soc., 2009, 131, 13498.
7 J. W. Barnes and G. D. Dorough, J. Am. Chem. Soc., 1950, 72, 4045.
¨
8 J. W. Buchler, J. Hu¨ttermann and J. Loffler, Bull. Chem. Soc. Jpn.,
1988, 61, 71.
9 J. A. Cissell, T. P. Vaid and G. P. A. Yap, J. Am. Chem. Soc., 2007, 129, 7841. 23 A. Ramasubramanian, V. Yurkiv, T. Foroozan, M. Ragone, R. Shahbazian-
10 A. Weiss, M. C. Hodgson, P. D. W. Boyd, W. Siebert and P. J.
Brothers, Chem. – Eur. J., 2007, 13, 5982.
11 J. Arnold, J. Chem. Soc., Chem. Commun., 1990, 976.
12 D. Y. Dawson and J. Arnold, J. Porphyrins Phthalocyanines, 1997, 1, 121.
13 J. Arnold, D. Y. Dawson and C. G. Hoffman, J. Am. Chem. Soc., 1993,
115, 2707.
Yassar and F. Mashayek, J. Phys. Chem. C, 2019, 123, 10237.
24 X.-B. Cheng, R. Zhang, C.-Z. Zhao, F. Wei, J.-G. Zhang and Q. Zhang,
Adv. Sci., 2016, 3, 1500213.
25 P. Gao, Z. Chen, Z. Zhao-Karger, J. E. Mueller, C. Jung, S. Klyatskaya,
T. Diemant, O. Fuhr, T. Jacob, R. J. Behm, M. Ruben and
M. Fichtner, Angew. Chem., Int. Ed., 2017, 56, 10341.
14 H. Brand, J. A. Capriotti and J. Arnold, Inorg. Chem., 1994, 33, 4334. 26 A. Stefanov, A. Stibor, A. Dominguez-Clarimon and M. Arndt,
15 M. Chen, H. Zhou, B. P. Klein, M. Zugermeier, C. K. Krug, H. J. Drescher, J. Chem. Phys., 2004, 121, 6935.
M. Gorgoi, M. Schmid and J. M. Gottfried, Phys. Chem. Chem. Phys., 2016, 27 P. Jakob and D. Menzel, J. Chem. Phys., 1996, 105, 3838.
18, 30643.
28 C. Weiland, A. K. Rumaiz, P. Pianetta and J. C. Woicik, J. Vac. Sci.
Technol., A, 2016, 34, 030801.
16 M. Schmid, S. R. Kachel, B. P. Klein, N. Bock, P. Mu¨ller, R. Riedel,
N. Hampp and J. M. Gottfried, J. Phys.: Condens. Matter, 2019, 31, 29 W. H. Gries, Surf. Interface Anal., 1996, 24, 38–50.
094002.
30 F. Buchner, K. Flechtner, Y. Bai, E. Zillner, I. Kellner, H. P.
Steinru¨ck, H. Marbach and J. M. Gottfried, J. Phys. Chem. C, 2008,
112, 15458.
17 T. E. Shubina, H. Marbach, K. Flechtner, A. Kretschmann, N. Jux,
F. Buchner, H. P. Steinru¨ck, T. Clark and J. M. Gottfried, J. Am.
Chem. Soc., 2007, 129, 9476.
31 M. W. Chase, Jr., J. Phys. Chem. Ref. Data, Monogr., 1998, 9, 1.
18 K. Diller, A. C. Papageorgiou, F. Klappenberger, F. Allegretti, J. V. Barth 32 D.-M. Chen, X. Liu, T.-J. He and F.-C. Liu, Chem. Phys. Lett., 2002,
¨
and W. Auwarter, Chem. Soc. Rev., 2016, 45, 1629.
361, 106.
13668 | Chem. Commun., 2019, 55, 13665--13668
This journal is ©The Royal Society of Chemistry 2019