COF-1: the first covalent organic framework

COF-1 needs only one building block, BDBA (benzene-1,4-diboronic acid): a benzene ring with a boronic acid, B(OH)₂, at each end. Three boronic acids lose three water molecules and close a flat six-membered boroxine ring, B₃O₃, of alternating boron and oxygen atoms.

Each boroxine ring joins three phenylene rods and each rod joins two boroxine rings, so each sheet is a honeycomb (hcb): B₃O₃ rings at the corners of hexagons, phenylene along their edges. The sheets are flat, about 15.65 Å from one hexagon centre to the next (the cell edge in its crystal file), and contain only boron, carbon, oxygen and hydrogen.

Unlike COF-5, the layers of COF-1 stack staggered (AB), in space group P6₃/mmc, about 3.35 Å apart (half of c ≈ 6.70 Å in its crystal file): every second layer is shifted so that its atoms sit over the pores of the layer below. This partly closes the channels, leaving pores of about 7 Å. As made, the pores hold mesitylene from the reaction solvent, giving the formula (C₃H₂BO)₆·(C₉H₁₂)₁.

1. Self-condensation. BDBA is suspended in a mixture of mesitylene and dioxane in a Pyrex tube, which is sealed and heated (solvothermal synthesis) (Côté et al. 2005). Three boronic acids close each boroxine ring, releasing water.
2. Self-correction. Boroxine formation is reversible, so B–O bonds form and break while the solid grows; misplaced molecules can come off again and the product is crystalline.
3. Activation. The mesitylene held in the pores is removed before gas sorption, and the evacuated framework is used for surface area and uptake measurements. Boroxine rings react with water (DeBlase et al. 2013), so the material is kept dry.

COF-1 was the first crystalline, porous solid built entirely from light elements joined by strong covalent bonds. Before 2005, covalent organic polymers were amorphous; Côté et al. used a reversible reaction, boroxine formation, so that errors could heal during growth and the product could be solved by powder X-ray diffraction. Its crystal structure is “entirely held by strong bonds between B, C, and O atoms” (Côté et al. 2005).

Together with COF-5, it opened the field of covalent organic frameworks and extended reticular chemistry from metal–organic frameworks to purely organic ones (Diercks & Yaghi 2017).

• Gas storage. COF-1 was one of seven COFs tested for hydrogen, methane and carbon dioxide storage; with its small pores it takes up less than COF-5 or the 3D COF-102 (Furukawa & Yaghi 2009).
• Stacking studies. Its staggered layers are the textbook contrast to eclipsed COF-5, showing how stacking sets the open pore size.
• Teaching. One monomer, one reaction and one ring make it the simplest example of how a COF is designed.
• Limitation. Boroxines hydrolyse; more water-stable linkages such as imines and β-ketoenamines (TpPa-1) were developed later (DeBlase et al. 2013; Kandambeth et al. 2012).

For a paper, slide or poster, show BDBA, the boroxine ring it forms, and two layers stacked staggered in two colours so the offset shows. The figure above has all of these and opens as an editable LabCanvas figure. The crystal structure drawing is a high-resolution SVG from the COF-1 crystal file and opens in the LabCanvas figure editor, where you can change the style, the number of cells and the view. The 3D model opens in the LabCanvas 3D Studio, where you can switch the stacking between staggered and eclipsed, add layers and export a high-resolution image.