Tetraethynylmethane: A Quest for the “Dutch Diamond”

Abstract

Tetraethynylmethane is an elusive building block that has motivated generations of organic chemists to provide an efficient synthesis. Despite being thermally unstable, several reports have culminated in its effective synthesis, and initial applications as porous material were described. However, the uniqueness of this molecule has not reached its full potential, both from a synthetic and applied point of view.

Keywords

Alkyne - Synthesis - Symmetric - Porous materials - Tetraethynylmethane

Since the discovery of fullerene in 1985, the field of synthetic carbon allotropes has rapidly expanded and led to various 2D and 3D allotropes of the typical carbon conformations, including various hybridizations.[1] The search for synthetic approaches to postulated small and large structures stimulated organic chemists to develop new methods.[2] The three-dimensional nature of the carbon atom stimulated research groups to develop expanded derivatives. Besides tetraphenylmethane,[3] the use of alkynes instead of para-substituted phenyl groups is intriguing. However, often tremendous synthetic difficulties were identified along the way, in return leading to new selective synthetic methods. The vinyl counterpart was significantly easier to obtain, and the first successful synthesis was reported in 1964 by Zimmermann.[4] Back in 1990, Alberts proposed the use of tetraethynylmethane (1) as a building block for diamondoid structures. Its T _d symmetry is fascinating and would lead to extended tetrahedral products. Due to its Dutch “windmill” shape, when drawn in 2D, he coined the respective 3D network a “Dutch Diamond.”[5] In addition, alternating alkynes and C(sp³)-carbon atoms lead to an interesting electron situation, where the extended π-bonds of the alkyne are in close proximity to one another.

Initial approaches toward the synthesis of tetraethynylmethane (1), however, failed and mostly led to triethynylmethane targets.[6] Extended tetraalkynylmethane derivatives were obtained, but feature additional flexible methylene groups, which in return provide lower rigidity.[7] With such extended molecules at hand, the formation of alkyne-incorporated and alkyne-extended “carbomers” of pyridine or cyclopentadienyl anions was proposed.[8] X-ray single-crystal analysis of the tris-TMS-protected tetraethynylmethane 2 revealed lengthened C(sp³)–C(sp) bonds and contracted C(sp)–C(sp) bonds, all deviating by around 0.04 Å from the expected values ([Fig. 1]).[9] Computational prediction on the DZP SCF level of theory for 2 confirmed this deviation in particular for the alkyne triple bonds (Δd = 0.039–0.046 Å).[10] Likewise, semi-empirical methods like PM3, AM1, and MNDO also tend to underestimate the C(sp³)–C(sp) and overestimate the C(sp)–C(sp) bonds in 1.[11] The use of density functional theory (DFT) on the B3LYP/6-311G** level indicated a much better agreement with experimentally obtained bond lengths and angles.[12] It was also observed that tetraethynylmethane (1) shows the largest HOMO-LUMO gap of 8.23 eV compared to its heavier tetrel analogues containing silicon and germanium. The experimentally found π-electron contraction of the C(sp)–C(sp) bonds in the crystal structure was rationalized by the high mobility of sp-hybridized electrons, which could be pushed outward due to electronic interactions of the four vicinal triple bonds.[10] Due to the nature of X-ray diffraction interacting with electron density, the use of neutron scattering was recommended to obtain a better picture of the bonding situation of 1 and 2.

Fig. 1 Molecular structure of tetraethynylmethane (1) and its tris-TMS-protected derivative 2 with the respective bond lengths obtained from X-ray single-crystal analysis (black, averaged for TMS-alkynes) or DZP SCF level of theory (gray).[9] [10]

Hence, future research is still required to better understand the unique bonding situation within tetraethynylmethane derivatives.

Since 1 is unstable at ambient conditions in the solid state, only theoretical bond lengths are available. The predicted heat of formation decreased from 236 to 76 kcal mol⁻¹ from 1 to 2, due to hyperconjugation effects of the TMS groups. The gas phase enthalpy of formation (ΔH°) for 1 was confirmed to be 242 kcal mol⁻¹ on the Gaussian-4 (G4) level of theory.[13] The strain energy of 1 was predicted to be only 4.8 kcal mol⁻¹, which is low and indicates little repulsion between the four alkynes.[10] The weak C–H⋯π interactions of solid tetraethynylmethane (1) were predicted to be −4.35 kcal mol⁻¹ on the PM3 semi-empirical level of theory, which could be a reason for its instability at room temperature.[14] Despite the missing structural proof, in the trace analysis of Kerogen combustion products, by time-of-flight secondary ion mass spectrometry (TOF-SIMS), the formation of C₉H₄ ^•+ or C₉H₅ ⁺, as well as their anionic derivatives, was confirmed.[15]

It was not until 1993 that the first successful synthesis of 1 was reported by the Feldman group.[9] Starting from 1,5-bis(trimethylsilyl)-1,4-pentadiyn-3-on (3), which can be obtained from the respective TMS-acetylene in two steps,[16] additional ten steps were necessary to obtain 19 mg of 1 in 16% overall yield ([Scheme 1]). The sequence included a Corey–Fuchs-type homologation and a Claisen rearrangement. Despite the deserved credit for the first successful report, the synthesis route needed rather forcing conditions, including the use of butyl lithium, LDA, or mCPBA, respectively. Shortly after, the same group reported an updated synthesis of tetraethynylmethane (1).[17] Like before, starting from compound 3, followed by ten steps, gave 19 mg of the product in an improved 30% overall yield. Further developments resulted in more carbon allotropes, based on intermediates of the tetraethynylmethane route.[18]

Scheme 1 Pioneering synthesis of tetraethynylmethane (1) by the group of Feldman.[9] [17]

For a bit more than a decade, new syntheses toward 1 remained elusive until the Yamaguchi group developed a new strategy, which streamlined the previous efforts ([Scheme 2A]).[19] The authors found that methylene-bridged bisalkynes undergo C−H activation reactions in the presence of gallium(III) chloride (GaCl₃) with haloalkynes. Out of a larger set of functionalized tetraethynylmethanes, they report the synthesis of tetrakis(triethylsilylalkynyl)methane (4a) in 22% and 56% yield, respectively, in just a single step from the respective precursors. In this coupling reaction, the bisalkyne 5a is reacted with TES-protected chloroacetylene (6) in the presence of 2 equiv of GaCl₃, 2 equiv of a bulky pyridine, and down to catalytic amounts of tert-butyldiphenylsilanol (7) in ortho-dichlorobenzene and methylcyclohexane as solvents at 150 °C. The fully TES-protected product 4a was obtained in 56% yield and 160 mg quantity, increasing by a factor of eight compared to Feldman’s work.[17] Later on, the same group achieved the functionalization of TES-protected propyne 8, among others, in a threefold coupling reaction, again utilizing GaCl₃ as a reagent ([Scheme 2B]).[20] Unfortunately, 4a was only obtained in 18% yield after 24 h of heating to 150 °C. However, this method represents a significant synthetic shortcut compared to the initial 10-step synthesis. Nonetheless, follow-up reports are scarce, and we were wondering about the synthetic access to 1.

Scheme 2 Improved and direct synthesis of 3a by Yamaguchi.[19] [20] cyMe: Methylcyclohexane, PhCl₂: 1,2-Dichlorobenzene.

Intrigued by the rapid access and fascinating structure of tetraethynylmethane (1), we were interested in applying this molecule as a building block in advanced materials. Since the Yamaguchi methods represent the most direct way,[19] [20] we envisioned the synthesis starting from bis(TES)pent-1,4-diyne 5a ([Scheme 3]). The substrate was obtained from TES-bromopropyne 9, synthesized from propagyl bromide (10), lithium hexamethyldisilazane (HMDS), and triethylsilyl chloride (TESCl) in 69% yield on a 4 g scale, which was then treated with TES-acetylene (11) and methyl-Grignard in a Copper-catalyzed reaction to obtain 1.35 g of 5a in 54% yield. Operational procedures were adapted from the literature.[21] TES-chloro acetylene (6) was obtained from TES-acetylene (11) with n-butyl lithium and N-chloro succinimide (NCS) in 49% yield and 1.7 g. tert-Butyldiphenylsilanol (7) was prepared from the hydrolysis of the chlorosilane 12 and gave 4.3 g in 93% yield. We then attempted the synthesis of 4, following the reported conditions[21] and were surprised to see no conversion. However, with freshly distilled starting materials 6 and 5a, and using a wider Schlenk flask, we successfully obtained TES-protected tetraethynylmethane 4a in 58-59% yield in quantities up to 170 mg.

Unfortunately, the same Gallium(III)-catalyzed or -mediated coupling reaction was unsuccessful with both bis(TMS)pent-1,4-diyne 5b and TES-propyne 14 ([Scheme 4]). Similarly, 5b was obtained by a cupration reaction of commercial 13, but was then unsuccessfully applied to the standard Gallium-mediated reaction with 6 to obtain mixed-protected 4b. The TMS groups seemed to be unstable under these conditions, as no TMS peaks were found in the ¹H NMR spectra of the reaction mixture. In addition, commercial 11 was methylated, and 1.1 g of 14 was obtained with n-BuLi and methyl iodide in 74% yield. Threefold gallium-catalyzed conversion of 14 toward 4a was again unsuccessful, and no product formation was observed.[20] Overall, we concluded that the twofold coupling reaction to obtain 4a was more reliable in our hands. Investigating the effect of other protection groups should be considered in the future to tune the reactivity of the alkyne intermediates.

Scheme 3 Synthesis of 4a and gram-scale access to its starting materials.

Scheme 4 Additional synthetic attempts toward 4a and 4b.

Given the limited availability of synthetic methods and the operational challenges associated with them, the synthetic organic community is requested to develop practical new methods. Such methods could capitalize on modern technologies employing electrosynthetic or photocatalytic activation, or mechanochemistry and ball milling, hence allowing for mild conditions and more benign starting materials.

Since tetraethynylmethane (1) quickly starts to decompose at ambient conditions within minutes, in situ deprotection and immediate follow-up transformations are necessary. With the target molecule in hand, initially, the group of Feldman investigated the use of tris-TMS-protected tetraethynylmethane 2 in a Glaser-Hay or Sonogashira coupling reaction, preceded or followed by TMS-deprotection with K₂CO₃ in MeOH to obtain the first products (e.g., 15) from 1 in good yields ([Scheme 5]).[17] Additional cross-coupling or derivatization products have not been reported to date.

Scheme 5 First cross-coupling example of in situ generated 1 by Feldman.[17]

With the optimized procedure for tris- or tetrakis-TMS-protected tetraethynylmethane 2 or 4b in hand, it was again the Feldman group to investigate the Glaser coupling of 1 to obtain alkyne−alkyne coupled oligomers.[22] Initial success gave small oligomers with up to 21 carbon atoms, corresponding to roughly three repeating units. More recently, the group of Valtchev utilized building block 4a to perform a copper-mediated Eglington coupling to yield an extended carbon network 17 from alkyne−alkyne coupling ([Fig. 2]).[21] The obtained black powder was applied as porous anode material for lithium-ion batteries and showed a high total capacity of around 780 mAh g⁻¹. Further applications in battery materials were reported.[23] Porous carbon networks with different alkyne lengths were investigated computationally for their lithium-ion uptake. Although having the smallest pores and pore volume, the “Dutch Diamond” material with only one alkyne between the C(sp³) atoms shows the highest density of 1.023 g cm⁻³ and the highest average reversible microporous capacity of 1200 mAh g⁻¹.[24] Nonetheless, the material shows slow charging kinetics, which makes it less ideal compared to its extended counterparts. Hence, up to this point, no synthetic access to the “Dutch Diamond” three-dimensional material was achieved, due to the alkyne dimerization reactivity of copper-catalyzed reactions.

More recently, the synthesis of the desired “Dutch Diamond” structure was achieved by the group of Zhang, who created 3D porous C(sp)-rich carbon material 16 by the reaction of bis(trimethylsilyl)acetylene and tetrabromomethane in the presence of fluoride.[25] The obtained material was porous with a BET surface area of 453 m² g⁻¹ and could be functionalized with rhodium to yield a hydroformylation catalyst with the highest conversion and turn-over frequency (TOF = 1558) among the studied catalysts. Given the condensation reaction conditions, it is to be expected that no well-defined polymeric network was obtained, but rather, polydisperse material was present. Experimental verification of the final “Dutch Diamond” network remains elusive to this point.

Fig. 2 Applications of the “Dutch Diamond” and alkyne-expanded derivatives as porous carbon materials (A). Inset adapted and reproduced with permission from Ref. [21]. One-step synthesis of the desired “Dutch Diamond” carbon network 16 (B).

Overall, it is to conclude that the quest for the “Dutch Diamond” is by far not over yet. New synthetic strategies are required in order to obtain large quantities of tetraethynylmethane building blocks from readily available starting materials. Since protection with TMS or TES is crucial to obtain bench-stable forms of 1, other protection groups should be investigated to provide a broader library of available molecules for follow-up derivatization. With such libraries in hand, cross-coupling reactions toward products like 15 would be highly interesting and would expand the field of synthetic carbon allotropes with distinct functions and properties. From such reactions, active probes with photophysical or redox properties can be incorporated. So far, applications of tetraethynylmethane derivatives have focused on battery materials, which is obviously an important field of application for low-weight and highly porous carbon allotropes, but disregards many other areas, like gas storage or separation, aviation materials, and many more. With many more building blocks in large quantities in hand, the quest for the “Dutch Diamond” and its many possible applications has really just begun.

Sebastian B. Beil

Sebastian studied chemistry at the CAU Kiel, Germany, and at the University of Stockholm. His research on hypervalent iodine and semiconducting polymers was performed with Prof. Berit Olofsson and Prof. Anne Staubitz. For his PhD, he moved to Mainz and graduated in 2019 under the supervision of Prof. Siegfried Waldvogel on electro-organic transformations. As an internship student, he spent time with Prof. Phil Baran at Scripps Research in La Jolla, California. After a first postdoc in the group of Prof. Max von Delius in Ulm working on synthetic carbon allotropes, he moved to Princeton University to work in the group of Prof. David MacMillan to develop new photoredox transformations. From 2021 to 2024, Sebastian was Assistant Professor in the Stratingh Institute for Chemistry at the University of Groningen, and since March 2024, he has held a research group leader position at the Max Planck Institute for Chemical Energy Conversion.

Conflict of Interest

The authors declare that they have no conflict of interest.

Acknowledgement

Dr. Jonas Hoffmann is acknowledged for his successful synthetic effort in reproducing and obtaining product 4a.[21] The graphical abstract was initially generated using Openart AI and subsequently modified by Dr. Sebastian Beil.

• References

• 1 Hirsch A. Nat Mater 2010; 9: 868-871
• 2 Diederich F, Rubin Y. Angew Chem Int Ed Engl 1992; 31: 1101-1123
• 3 Chen L, Huo J, Zhang J. et al. Green synth Catal 2023; 4: 273-292
• 4 Berger JG, Stogryn EL, Zimmerman AA. J Org Chem 1964; 29: 950-951
• 5 Alberts AH. In Lemstra PJ, Kleintjens LA. eds Integration of Fundamental Polymer Science and Technology—4. Dordrecht, Netherlands: Springer; 1990: 396-401
• 6a Alberts AH, Wynberg H, Chem J. Soc, Chem Commun 1988; 748-749
• 6b Alberts AH. J Am Chem Soc 1989; 111: 3093-3094
• 7 Bunz U, Vollhardt KPC, Ho JS. Angew Chem Int Ed Engl 1992; 31: 1648-1651
• 8 Chauvin R. Tetrahedron Lett 1995; 36: 397-400
• 9 Feldman KS, Kraebel CM, Parvez M. J Am Chem Soc 1993; 115: 3846-3847
• 10 Ma B, Sulzbach HM, Xie Y, Schaefer III HF. J Am Chem Soc 1994; 116: 3529-3538
• 11 Capitani JF. J Mol Struct (THEOCHEM) 1995; 332: 21-23
• 12 Tatevosyan MM, Zhukova TN, Vlasenko VG. J Struct Chem 2021; 62: 1684-1693
• 13 Rayne S, Forest K. Comput Theor Chem 2011; 970: 15-22
• 14 Raynor S, Song HH. ACS Omega 2020; 5: 27546-27555
• 15 Sjövall P, Bake KD, Pomerantz AE, Lu X, Mitra-Kirtley S, Mullins OC. Fuel 2021; 286: 119373
• 16a Müller E, Segnitz A. Liebigs Ann Chem 1973; 1973: 1583-1591
• 16b Hauptmann H. Tetrahedron 1976; 32: 1293-1297
• 16c Posner GH, Loomis GL, Sawaya HS. Tetrahedron Lett 1975; 16: 1373-1376
• 17 Feldman KS, Weinreb CK, Youngs WJ, Bradshaw JD. J Am Chem Soc 1994; 116: 9019-9026
• 18 Feldman KS, Weinreb CK, Youngs WJ, Bradshaw JD. J Org Chem 1994; 59: 1213-1215
• 19 Amemiya R, Suwa K, Toriyama J, Nishimura Y, Yamaguchi M. J Am Chem Soc 2005; 127: 8252-8253
• 20 Amemiya R, Yamaguchi M. Adv Synth Catal 2007; 349: 1011-1014
• 21 Zhao Z, Das S, Xing G. et al. Angew Chem Int Ed 2018; 57: 11952-11956
• 22 Feldman KS, Mareska DA, Weinreb CK, Chasmawala M. Chem Commun 1996; 865-866
• 23 Huang CY, Song Y. CN108892119 2018
• 24 Heasman P, Varley E, Trewin A. Energy Fuel 2022; 36: 6560-6568
• 25 Wang J, Fu X, Yan N, Zhang Y. ChemSusChem 2021; 14: 3806-3809

Publication History

Received: 08 June 2025

Accepted after revision: 14 July 2025

Accepted Manuscript online:
14 July 2025

Article published online:
04 September 2025