Solvatochromic 4-(2,4,6-triphenylpyridinium-1-yl)phenolate

In this write-up the synthesis of 4-(2,4,6-triphenylpyridinium-1-yl)phenolate is described. This solvatochromic betain dye is related to the more commonly known "Reichardt's dye" by lacking the 2,6-diphenyl substitution on the phenolate. While this change substantially alters the color profile (and probably solubility), this compound is much more accessible.

This synthesis uses triphenylpyrylium bisulfate (though other triphenylpyrylium salts should substitute) which was made previously (a writeup is in the works).



p-Aminophenol from paracetamol
The procedure was adapted from the one shown by chemplayer.^[1] Modifications include increasing the reaction time from one hour to two hours, the use of argon, and the addition of sodium dithionite to reduce oxidation. The yield was comparable, but the product was less brown.

Synthesis:
3.0 g (20 mmol) Paracetamol (extracted from pills with ethanol and recrystallized from boiling water), 3.0 g (> 90 %, ca. 48 mmol) KOH flakes, 0.2 g (ca. 80 %, ca. 1 mmol) sodium dithionite, 20 g of ice and 30 mL of water was placed together with a stir bar in a 100 mL flat-bottom flask. Then a reflux condenser was placed on top and the headspace and the condenser were flushed with argon. The top was loosely covered to limit air exchange, but ideally an argon bag would be used, as I had to find out later. The solution was refluxed in subdued light for two hours, during which time it stayed completely colorless. On cooling air sucked back in and the surface layer immediately became brown, hence a bag of argon should be used to prevent this. The solution was cooled in a water bath while a slow stream of argon was passed through it and dilute sulfuric acid (ca. 7.5 g, ca. 11 % in water, ca. 8 mmol) was used to bring the pH to 4-5, a white to tan precipitate started to form and the solution was stirred for one hour on ice. It was filtered over a Büchner funnel with vacuum and washed with little ice cold water, then it was air dried in the dark.
Yield: 1.35 g (62 %)

The filtrate was pH-readjusted as it had become basic again, saturated with sodium chloride, placed in the fridge for a week, then the additional precipitate was filtered off, washed, and dried in the dark.
An additional 0.56 g was obtained.
Total yield 1.91 g (88 %).

The product was very slightly tan or cream colored, just a little bit off-white. It did slowly turn brown during the next weeks, so presumably it would be best to also store it under argon.


Fig. 1: p-Aminophenol.

Analysis:
TLC was performed on silica 60 F₂₅₄ with 2/1 EtOAc/n-hexane showed paracetamol at an R_f of 0.21 and the product at R_f 0.36 with no visible paracetamol contamination under 270 nm UV or with Hanessians stain (prepared with ammonium heptamolybdate, ceric ammonium sulfate and sulfuric acid as per recipe 2 on curly arrow^[2]).


Fig. 2: TLC of paracetamol (left) and p-aminophenol (right) with co-spot (middle).

4-(2,4,6-Triphenylpyridinium-1-yl)phenolate hexahydrate
The literature procedure^[3] was followed, but the triphenylpyrylium iodide was substituted with triphenylpyrylium bisulfate.

Synthesis:
Into a 100 mL round bottom flask were weighed, in this order: 1.00 g (2.5 mmol, 1 eq.) 2,4,6-triphenylpyrylium bisulfate, 0.40 g (4.9 mmol, 2 eq.) anhydrous sodium acetate, 0.40 g (3.7 mmol, 1.5 eq.) p-aminophenol and 25 mL glacial acetic acid together with a stir bar. A reflux condenser was attached and the mix refluxed for about fifteen minutes directly on the hotplate. The suspension turned reddish during this time. Afterwards the warm solution was dumped onto some crushed ice.
Then, 35 g (620 mmol) KOH were dissolved in 50-60 mL of ice water and slowly added. The yellow precipitate turned into reddish solution and then into a reddish-brown suspension. Almost all ice melts. Stirring was continued for another hour and then filtered with vacuum on a Büchner funnel. The precipitate was washed with a little KOH solution and is fine, somewhat pasty and inhomogenously lumpy red-brown.


Fig. 3-6: Progression of the reaction.

The precipitate was then recrystallized by dissolving it in 100-120 mL boiling 5 % KOH (about 2 g KOH per 100 mL water) and filtered hot through filter paper (a lot remains in the filter paper, keep it). Directly on cooling about 3 mm long, red glittering needle shaped crystals began forming in the red solution. The next day the crystals were filtered off and the residue from the hot filtration was recrystallized from this filtrate in the same manner.

Yield: 0.880 g (1.7 mmol, 70 %) of the betain hexahydrate.


Fig. 7 and 8: Recrystallization and product.

A few crystals were placed in a beaker and carefully dehydrated with a heatgun set to 190 °C. This dried product is dark blue. (The literature dries it with phosphorous pentoxide under vacuum at 140-150 °C.)

Solvatochromism:
Solutions of the dehydrated betain are strongly solvatochromic and somewhat thermochromic.

The colors are, in order of their appearance (Solvent - Cold / Hot):
EtOAc - yellow like chromate / yellow like chromate
EtOH - orange / red-orange
DMSO - red-brown / blue-purple
Acetone - purple / blue-purple
THF - dark grey-blue / somewhat more greenish lighter blue
DCM - dark blue / (not tested)
Xylene - wasn't soluble


Fig. 9: Solvatochromicity of the dehydrated betain. From left to right: EtOAc, EtOH, DMSO, acetone, THF, DCM.

Analysis:
TLC with 4/6 EtOAc/n-hexane gave no moving spots on the TLC plate, suggesting the absense of p-aminophenol. Redevolpment of the same plate with the dehydrated product in 2/1 DCM/EtOH gave spots at R_f 0.11 and 0.56 under 250 nm UV. Vanillin stain did not visualize anything. This will have to be redone properly. However I believe the solvatochromicity and high crystallinity of the product indicate identity and probably good quality, for the time being.

Literature:
[1] - ChemPlayer, "p-Aminophenol preparation from paracetamol", BitChute 2018, https://www.bitchute.com/video/FluX5CkOEiLp (reupload from deleted youtube channel)
[2] - D. Sejer, "Let's talk about TLCs Part 2 - Hanessian's Stain", Curly Arrow 2006, http://curlyarrow.blogspot.com/2006/11/lets-talk-about-tlcs-...
[3] - W. Schneider, W. Döbling, R. Cordua, "Über Mesomerie bei N-Oxyphenyl-pyridinium-Basen.", Ber. dtsch. Chem. Ges. A/B 1937, 70, 1645-1665, https://doi.org/10.1002/cber.19370700802