Open Access Article
Jack M. Hemingway*a,
Xiao Yangb,
Paul G. Waddell
a,
Joel Corneliob,
Matthew E. Lowea,
James A. Dawson
a,
Peter R. Slater
bc,
Roly J. Armstrong
a and
Erli Lu
*b
aChemistry-School of Natural and Environmental Sciences, Newcastle University, Newcastle upon Tyne, NE1 7RU, UK. E-mail: jack.hemingway@newcastle.ac.uk
bSchool of Chemistry, University of Birmingham, Birmingham, B15 2TT, UK. E-mail: e.lu@bham.ac.uk
cThe Faraday Institution, Quad One, Becquerel Avenue, Harwell campus, Didcot OX11 0RA, UK
First published on 25th November 2025
Separation of lithium and sodium is a topic of substantial scientific and industrial importance. Regarding Li/Na perchlorates, which are not only environmental hazards but also useful oxidants in chemical synthesis, an efficient Li–Na perchlorate separation method has not been reported due to lack of a ligand which can selectively coordinate with one of the two. Herein, we report an efficient Li–Na perchlorate separation by using our hexadentate ligand N,N′,N″-tris-(2-N-diethylaminoethyl)-1,4,7-triaza-cyclononane (DETAN), which can selectively coordinate with LiClO4 at room temperature to form a monomer in excellent yield but does not coordinate with NaClO4 even at elevated temperature. The structure of the monomeric complex, [LiClO4(DETAN)] (1), was characterised by single-crystal X-ray diffraction and NMR spectroscopy.
Understandably, the anion also influences the Li/Na separation by intervening with the ligand coordination to the metal cation. As such, usually a ligand system which works for one anion (such as the Li/NaNTf2 in the aforementioned Nitschke/Wales system11) cannot be directly translated to another anion. In this work, we focus on the perchlorate anion (ClO4−). Featuring a tetrahedral structure with distributed charge at four O atoms,12 perchlorate is a potential public health concern due to its toxicity.13 Also, lithium perchlorate (LiClO4) is of interest in organic synthesis (e.g. accelerating Diels–Alder reactions14 and promoting cyanosilylation of carbonyl compounds15) and as an electrolyte component in Li-ion batteries.16 In the natural environment, LiClO4 usually co-exists with its heavier group-1 sister, sodium perchlorate (NaClO4), and their sensing and isolation has been of long-standing scientific interest since the early 20th century.17 Yet, there is the lack of an efficient ligand that can separate LiClO4 and NaClO4. In this work, we fill the knowledge gap by using a hexadentate N-donor ligand N,N′,N″-tris-(2-N-diethylaminoethyl)-1,4,7-triaza-cyclononane (DETAN), which was developed by us in 2021.18 The DETAN ligand has exhibited versatile coordination features with group-1 alkali metal cations.19,20 Herein we describe the selective binding of DETAN to LiClO4 in a 1
:
1 molar mixture of Li/NaClO4, which results in the separation of the two perchlorates. The DETAN–LiClO4 complex is proven to be a monomer by single-crystal X-ray diffraction (SCXRD), and its electronic structure is studied using DFT computations. The full details are elaborated on in the following sections.
We would like to bring to our readers’ awareness that this Article focuses on ligand design and aims at understanding the coordination behaviours of the DETAN ligand towards LiClO4 and NaClO4 in non-aqueous non-coordinative toluene solution. Water molecules, or any other coordinative solvents (such as THF), would profoundly change the coordination dynamic, hence the conclusions drawn here are not necessarily applicable in an aqueous, or coordinative solvent, environment.
:
1 molar ratio was monitored by both 1H and 7Li NMR. At room temperature within approximately 2 days, the starting material LiClO4 was consumed, with the concomitant formation of a new set of 1H NMR signals corresponding to a DETAN-coordinated species, as well as a new 7Li NMR signal at 0.19 ppm (see SI for the NMR spectra, Fig. S2). Scaling up the reaction in toluene at room temperature for 2 days and the following crystallisation led to the isolation of [Li(κ4-N-DETAN)(η1-O-ClO4)] (1) in 66% crystalline yield (Fig. 1a).
Single crystals of 1 suitable for SCXRD analysis were isolated from its diethyl ether solution, and its molecular structure is exhibited in Fig. 2. 1 is a monomeric LiClO4 complex, joining the number of previously reported examples of such structures.21 Though 1 is not the first LiClO4 monomer complex, it does have a few intriguing structural features. The DETAN ligand in 1 coordinates to the Li+ centre in a κ4 mode, with all the three nitrogen atoms in the macrocycle and one of the three sidearms coordinated. This DETAN coordination mode is similar to our previously reported DETAN-coordinated lithium iodide [Li(κ4-N-DETAN)(I)].19 The perchlorate anion in 1 coordinates to the Li+ centre in a η1 mode through one of its four oxygen atoms. The Li–O bond length is 2.063(3) Å, which is substantially shorter (by approximately 0.14 Å) than a structurally relevant monomeric LiClO4 complex ligated with a tripodal tripyridyl ligand TPA, [Li(κ4-N-TPA)(η1-O-ClO4)], where the Li–O bond length was reported at 2.2023(3) Å.21p The short Li–O bond in 1 cf. that in [Li(κ4-N-TPA)(η1-O-ClO4)] could be a result of the weaker N → Li dative bonds in 1 (avg. 2.16 to 2.25 Å) than in the TPA complex (2.08 to 2.18 Å), which renders the Li+ centre in 1 more positively charged and hence the shorter Li–O bond. The four Cl–O bonds in ClO4− can be divided into two groups: (1) the Cl1–O1 bond at 1.4476(10) Å; (2) the Cl1–O2/O3/O4 bonds at 1.418–1.428 Å. The former is longer than the latter, likely due to the coordination with Li+, which withdraws the electron density. Despite the small variation, all four Cl–O bond lengths are within the normal range for a ClO4− anion,21 and should be treated as double bonds.
Generally speaking, all the applications and hazards of the perchlorate anion are underpinned by its unique structure of a formally high-valent chlorine centre (positively charged) and four electron-rich oxygens (negatively charged). In this regard, a similarity can be drawn between the perchlorate anion and topical hypervalent organoiodine compounds, which are also predominantly used as oxidants.22 Indeed, close examination of the inherent electronic structure of ClO4− has inspired the design of an iron-catalyst system for perchlorate reduction.23,24 Given this context, in an effort to explore the electronic structure of 1, we conducted Density Functional Theory (DFT) calculations (see SI for details).
A natural population analysis was completed to probe the atomic charges in the structure of 1, the results of which are outlined in Fig. 3(a). Additionally, the electrostatic surface potential (ESP) was calculated and is outlined.
![]() | ||
| Fig. 3 The NPA charges of several atoms of interest in the structure of 1 (a) alongside the calculated electrostatic surface potential (b). | ||
The positive NPA charges are located on the Li and Cl and found to be +0.835e and +2.551e, respectively, whilst the negative charges are primarily located on the four O atoms, with O1 having the largest absolute charge of −0.970e when compared to O2, O3 and O4, which exhibit an average charge of −0.840e. The four nitrogen atoms that coordinate to the Li+ also exhibit negative charges (average charge of −0.483e). The ESP diagram also highlights the negative charge build up over the perchlorate species.
Since the DETAN ligand coordinates to Li+ in 1 in a κ4-N mode, and two of the three sidearms remain coordination-free, we explored the possibility of replacing the DETAN ligand with a typical κ4-N ligand, namely Me6Tren. In our previous reports, both the DETAN and the Me6Tren ligands were found to be able to coordinate to Li+, though their kinetic behaviours in solution are quite different.20 Surprisingly, in this case we find that the Me6Tren does not coordinate with either LiClO4 or NaClO4 even at elevated temperatures (60 °C) (Fig. 1). Though the reason(s) behind this somewhat surprising finding remain unclear, we hypothesise that the substantially different kinetic features between the fully flexible Me6Tren and the semi-rigid DETAN play an important role.20 This rigid nature was indeed a key part of our initial design concept of the DETAN ligand.18 In a very recent report, we also found that the different coordination kinetic features of Me6Tren vs. DETAN led to a pronounced difference in their performance in a ligand-promoted alkali metal silylalkyls mediated C
O bond methylenation.25
After the isolation of 1, we tested the DETAN coordination reaction with NaClO4 also in d6-benzene (Fig. 1b). Interestingly, there was no coordination between NaClO4 and DETAN even at 60 °C. It is obvious that the DETAN ligand exhibits highly selective coordination to LiClO4, but not NaClO4. This is the first ligand with such a selective coordination behaviour between Li/NaClO4.
:
1 molar mixture. A solution of DETAN in toluene was added to a well-ground solid mixture of LiClO4 and NaClO4 and stirred at room temperature for 2 days (Fig. 4). After the reaction, the residual solid phase and the organic phase were treated in vacuo to remove all volatiles and subsequently analysed by 7Li and 23Na NMR spectroscopy. In the organic phase, a strong singlet peak corresponding to complex 1 was observed in the 7Li NMR spectroscopy, while no sodium-containing species were detected in the 23Na NMR. In contrast, the solid phase (dissolved in D2O) exhibited a strong 23Na NMR signal corresponding to NaClO4 and only a very weak 7Li signal, confirming a minimal Li+ retention. Only Li and Na were determined in the organic and solid phases, respectively, by inductively coupled plasma optical emission spectrometry (ICP-OES). The ICP-OES concentrations suggested that the recovery yield of Li in the organic phase was 99%, with 70% of Na recovered in the solid phase (the slightly lower yield of NaClO4 is due to the filtration, where a small amount of NaClO4 was retained on the filter paper and could not be recovered). We optimised the filtration process and the recovery of Li and Na can reach 94% and 96%, respectively (see SI for separation experiment, batch 2 and Table S1). Additionally, we can also recycle approximately 68% DETAN ligand and recover 78% of Li through a following liquid–liquid back extraction using deionised water (see SI for back-extraction experiment, batch 5 and Table S2). The following test for the solubility of perchlorates (LiClO4 and NaClO4) showed that the perchlorates were not dissolved in both toluene and benzene (see SI for solubility test, batches 3 to 4 and Table S1).
In order explore the influence of the anionic component on the coordination of DETAN to the Li/Na salts, we selected an additional salt (tetrafluoroborate, BF4−) on the basis of its similar tetrahedral geometry and charge distribution. Once again, the DETAN ligand preferentially bound to the Li salt (LiBF4) over the Na salt, mirroring the selectivity in the case of the perchlorate anion (see SI for details regarding selectivity and solubility experiments, batches 6–8 and Table S3). This is particularly interesting as previous work focusing on iodide (I−) and tetraphenylborate (BPh4−) anions has shown that coordination utilising DETAN is achievable for both Li/Na salts, unlike ClO4− and BF4−.19
CCDC 2481385 (1) contains the supplementary crystallographic data for this paper.26
:
2 Salt : Base Ratio, Z. Anorg. Allg. Chem., 2006, 632, 1829–1838, DOI:10.1002/zaac.200600080;
(d) A. A. Bezdomnikov, V. E. Sharov, L. I. Demina, M. I. Skrebtsov, A. B. Ilyukhin and A. Y. Tsivadze, Specific features of lithium solvent extraction from perchlorate media with benzo-15-crown-5, Polyhedron, 2023, 244, 116612, DOI:10.1016/j.poly.2023.116612;
(e) R. R. Burnette and I. A. Guzei, CCDC 761188: Experimental Crystal Structure Determination. CSD Communication, 2016, DOI:10.5517/ccdc.csd.cctk2gw;
(f) H. Gohil, S. Chatterjee, S. Yadav, E. Suresh and A. R. Paital, An Ionophore for High Lithium Loading and Selective Capture from Brine, Inorg. Chem., 2019, 58, 7209–7219, DOI:10.1021/acs.inorgchem.9b00135;
(g) J. Dale, J. Eggestad, S. B. Fredriksen and P. Groth, 1,5,9-Trioxacyclododecane and 3,3,7,7,11,11-hexamethyl-1,5,9-trioxacyclododecane: novel lithium cation complexing agents, J. Chem. Soc., Chem. Commun., 1987, 1391–1393, 10.1039/C39870001391;
(h) K. Ueno, K. Hiratani, K. Taguchi, T. Okada and H. Sugihara, Crystal Structure of a 1
:
1 Complex of 1,3-Bis(8-quinolyloxy)propane and Lithium Perchlorate, Chem. Lett., 1987, 16, 949–952, DOI:10.1246/cl.1987.949;
(i) B. S. Bouley, F. Tang, D. Y. Bae and L. M. Mirica, C–H bond activation via concerted metalation–deprotonation at a palladium(III) center, Chem. Sci., 2023, 14, 3800–3808, 10.1039/D3SC00034F;
(j) G. W. Buchnan, R. A. Kirby and J. P. Charland, An example of conformational disorder in solids: X-ray crystallographic and 13C CPMAS NMR studies of dibenzo and dicyclohexano-13-crown-4 ethers and some lithium complexes, Can. J. Chem., 1990, 68, 49, DOI:10.1139/v90-011;
(k) D. M. Seo, T. Afroz, J. L. Allen, P. D. Boyle, P. C. Trulove, H. C. De Long and W. A. Henderson, Structural Interactions within Lithium Salt Solvates: Cyclic Carbonates and Esters, J. Phys. Chem. C, 2014, 118, 25884–25889, DOI:10.1021/jp5079168;
(l) T. N. Y. Hoang, M. Humbert-Droz, T. Dutronc, L. Guénée, C. Besnard and C. Piguet, A Polyaromatic Terdentate Binding Unit with Fused 5,6-Membered Chelates for Complexing s-, p-, d-, and f-Block Cations, Inorg. Chem., 2013, 52, 5570–5580, DOI:10.1021/ic400526j;
(m) J. J. Vittal, X. Wang and J. D. Ranford, Influence of the Li+ on the Structure of the [Cu3(phis)3]3+ Cation, Inorg. Chem., 2003, 42, 3390–3392, DOI:10.1021/ic026311n;
(n) A. Merz, T. Futterer, J. Lex and H. Inerpwicz, Alkali Metal Complexes of o, o′-(Tetraethyleneglycoldiyl)-(Z)-stilbene: One Common Ligand Conformation for Li, Na, K, and Rb Ions, Angew. Chem., Int. Ed. Engl., 1997, 36, 278–280, DOI:10.1002/anie.199702781;
(o) K. Ziach and J. Jurczak, Mirror symmetry breaking upon spontaneous crystallization from a dynamic combinatorial library of macrocyclic imines, Chem. Commun., 2015, 51, 4306–4039, 10.1039/C4CC10083B;
(p) N. G. Spiropulos, E. A. Standley, I. R. Shaw, B. L. Ingalls, B. Diebels, S. V. Krawczyk, B. F. Gherman, A. M. Arif and E. C. Brown, Synthesis of zinc and cadmium O-alkyl thiocarbonate and dithiocarbonate complexes and a cationic zinc hydrosulfide complex, Inorg. Chim. Acta, 2012, 386, 83–92, DOI:10.1016/j.ica.2012.01.040.| This journal is © the Partner Organisations 2025 |