Multicomponent reactions provide key molecules for secret communication

A convenient and inherently more secure communication channel for encoding messages via specifically designed molecular keys is introduced by combining advanced encryption standard cryptography with molecular steganography. The necessary molecular keys require large structural diversity, thus suggesting the application of multicomponent reactions. Herein, the Ugi four-component reaction of perfluorinated acids is utilized to establish an exemplary database consisting of 130 commercially available components. Considering all permutations, this combinatorial approach can unambiguously provide 500,000 molecular keys in only one synthetic procedure per key. The molecular keys are transferred nondigitally and concealed by either adsorption onto paper, coffee, tea or sugar as well as by dissolution in a perfume or in blood. Re-isolation and purification from these disguises is simplified by the perfluorinated sidechains of the molecular keys. High resolution tandem mass spectrometry can unequivocally determine the molecular structure and thus the identity of the key for a subsequent decryption of an encoded message.

Interestingly, a similar fragmentation pattern as for the tandem-MS spectra can be observed. e, Fragment assignment of the weak signal at 10.6 min indicates the presence of a Ugi product with a shorter perfluorinated side chain. This impurity (originating from a shorter perfluorinated acid) was already present in the starting material and did not interfere with other analytical methods and the readout.        Fast-atom-bombardment (FAB) and electron ionization (EI) spectra were recorded utilizing a Finnigan MAT 95 mass spectrometer. Molecule fragmentations observed in FAB or EI measurements were formally denoted as homolytic bond cleavage to allow a simple illustration of the observed m/z species, but a radical mechanism (or formation) was not proven.
Infrared (IR) spectra were recorded on a BRUKER Alpha-p instrument applying ATR-technology.
The signals were noted from large to smaller wavenumbers with the following notation: IR (Type of measurement)  [cm -1 ] = Wave number (signal intensity, molecular oscillation assignment). The signal shape and intensity is reported relative to the signal of highest intensity and was abbreviated in the following pattern: br = brought, vs = very strong, s = strong, m = medium, w = weak, vw = very weak.  1.18 (m,9 H,CH3 18,28,29 ), 1.14 -0.96 (m,4 H,CH2 19,20
Subsequently, cyclohexylisocyanide (79.7 µL, 70.0 mg, 641 µmol, 1.70 eq.) was added to the stirring mixture. The reaction was stirred for 3 d at room temperature. The crude reaction mixture was dried under reduced pressure and purified via column chromatography employing FluoroFlash ® silica gel.
The fluorous fraction was tested for purity via TLC and concentrated under reduced pressure. The remaining perfluorononanoic acid was removed with a short silica gel filter column, eluting with c-hexane/ethyl acetate (3:1) to yield the Ugi product as a yellow oil (41.8 mg, 52.8 µmol, 14.0%).
The solid was washed with 10 mL methanol three times. Subsequently, the filtrate was concentrated under reduced pressure. Perfluorononanoic acid (175 mg, 377 µmol, 1.00 eq.) dissolved in 1 mL methanol was added to the imine at room temperature and the resulting mixture was stirred for 2 min.
Subsequently, tert-butylisocyanide (51.2 µL, 37.6 mg, 453 µmol, 1.20 eq.) was added to the stirring mixture. The reaction was stirred for 3 d at room temperature. The crude reaction mixture was dried under reduced pressure and purified via column chromatography employing FluoroFlash ® silica gel.
The fluorous fraction was concentrated and the residue was adsorbed onto celite ® and purified via

Ugi reaction of perfluorononanoic acid, isobutyraldehyde, pentylisocyanide and cyclohexylamine
In a 25 mL round bottom flask isobutyraldehyde (46.2 mg, 641 µmol, 1.70 eq.) was dissolved in 1.5 mL methanol, subsequently cyclohexylamine (674 µL, 63.6 mg, 641 µmol, 1.70 eq.) was added and the resulting mixture was stirred for 60 min over sodium sulfate. Afterwards, the mixture was filtrated and the solid was washed with 10 mL methanol three times. Subsequently, the filtrate was concentrated under reduced pressure. Perfluorononanoic acid (175 mg, 377 µmol, 1.00 eq.) dissolved in 0.5 mL methanol was added to the imine at room temperature and the resulting mixture was stirred for 2 min. Subsequently, pentylisocyanide (80.6 µL, 62.2 mg, 641 µmol, 1.70 eq.) was added to the stirring mixture. The reaction was stirred for 6 d at room temperature. The crude reaction mixture was dried under reduced pressure and purified via column chromatography employing FluoroFlash ® silica gel. The fluorous fraction was tested for purity via TLC and concentrated under reduced pressure. The remaining perfluoro acid was removed with a short silica gel filter column, eluting with c-hexane/ethyl acetate (3:1) to yield the Ugi product as a highly viscous yellow oil (9.1 mg, 12.6 µmol, 3.34%).

Ugi reaction of perfluorononanoic acid, isobutyraldehyde, cyclohexylisocyanide and cyclohexylamine
In a 25 mL round bottom flask isobutyraldehyde (46.2 mg, 641 µmol, 1.70 eq.) was dissolved in 1.5 mL methanol, subsequently cyclohexylamine (73.5 µL, 63.6 mg, 641 µmol, 1.70 eq.) was added and the resulting mixture was stirred for 60 min over sodium sulfate. Afterwards, the mixture was filtrated and the solid was washed with 10 mL methanol three times. Subsequently, the filtrate was concentrated under reduced pressure. Perfluorononanoic acid (175 mg, 377 µmol, 1.00 eq.) dissolved in 0.5 mL methanol was added to the imine at room temperature and the resulting mixture was stirred for 2 min. Subsequently, cyclohexylisocyanide (79.9 µL, 70.0 mg, 641 µmol, 1.70 eq.) was added to the stirring mixture. The reaction was stirred for 6 d at room temperature. The crude reaction mixture was dried under reduced pressure and purified via column chromatography employing FluoroFlash ® silica gel. The fluorous fraction was tested for purity via TLC and concentrated under reduced pressure. The remaining perfluoro acid was removed with a short silica gel filter column, eluting with c-hexane/ethyl acetate (3:1) to yield the Ugi product as a highly viscous yellow oil (31.2 mg, 42.9 µmol, 11.4%).      1.22 (m,9 H,CH3 19,29,30

Ugi reaction of perfluorononanoic acid, heptanal, 4-methoxyphenylisocyanide and butylamine
In a 25 mL round bottom flask heptanal (71.0 µL, 56.0 mg, 490 µmol, 1.30 eq.) was dissolved in 1.5 mL methanol, subsequently butylamine (48.5 µL, 35.9 mg, 490 µmol, 1.30 eq.) was added and the resulting mixture was stirred for 60 min over sodium sulfate. Afterwards, the mixture was filtrated and the solid was washed with 10 mL methanol three times. Subsequently, the filtrate was concentrated under reduced pressure. Perfluorononanoic acid (175 mg, 377 µmol, 1.00 eq.) dissolved in 1 mL methanol was added to the imine at room temperature and the resulting mixture was stirred

Supplementary Figure 197 | HMBC experiment of the title compound recorded in CDCl3
Ugi reaction of perfluorononanoic acid, benzaldehyde, cyclohexylisocyanide and pentylamine In a 25 mL round bottom flask benzaldehyde (56.3 µL, 49.4 mg, 453 µmol, 1.20 eq.) was dissolved in 1.5 mL methanol, subsequently pentylamine (56.6 µL, 42.7 mg, 490 µmol, 1.30 eq.) was added and the resulting mixture was stirred for 60 min over sodium sulfate. Afterwards, the mixture was filtrated and the solid was washed with 10 mL methanol three times. Subsequently, the filtrate was concentrated under reduced pressure. Perfluorononanoic acid (175 mg, 377 µmol, 1.00 eq.) dissolved in 1 mL methanol was added to the imine at room temperature and the resulting mixture was stirred for 2 min. Subsequently, cyclohexylisocyanide (56.3 µL, 59.4 mg, 453 µmol, 1.20 eq.) was added to the stirring mixture. The reaction was stirred for 4 d at room temperature. The crude reaction mixture was dried under reduced pressure and purified via column chromatography employing FluoroFlash ® silica gel. The fluorous fraction was tested for purity via TLC and concentrated under reduced pressure. The remaining perfluoro acid was removed with a short silica gel filter column, eluting with c-hexane/ethyl acetate (3:1) to yield the Ugi product as a colorless solid (106 mg, 140 µmol, 42.3%).

Influence of stereochemistry
During the Ugi reaction a new chiral center is formed, which is not controlled in the present protocol.
However, if chiral precursor components are utilized, diastereomeric product mixtures will result. In order to study if diastereoisomers have an influence on the separation protocol or the MS/MS fragmentation behavior, a model substrate was synthesized, purified via F-SPE and fragmented via ESI-MS/MS.

Supplementary Figure 211
| Model substrate for studying the influence of stereochemistry. The structure presented above was synthesized from the racemic precursor components. The reaction product will thus be a mixture of four different diastereoisomers.
The F-SEP purification protocol retains perfluorinated compounds selectively and is thus unaffected by diastereoisomers. This is a great advantage of F-SPE and was utilized in the field of so called fluorous mixture synthesis (FMS). 2 FMS separation was also applied successfully for the synthesis of diastereomer mixtures, wherein F-tagged diastereomers were synthesized stepwise and separated from non-fluorinated contaminations via F-SPE. 3 The synthetic procedures for the diastereomeric molecular key mixture presented above is included at the end of the chapter Synthetic procedures (page 154). If a diastreomeric mixture is subjected to a tandem-MS experiment, it can be assumed that the required fragments for the readout will still be observed, since, even if theoretically different fragmentation pathways may occur, the favored fragments presented in the manuscript will still be observed, at least for some of the fragmentation pathways (maybe with a somewhat lower probability, but the required fragments should still be present and detectable). The observed mass of the fragments is independent from the stereochemical information. The fragmentation behavior of diastereoisomers in MS/MS methods and was previously studied, confirming that the fragmentation leads to a different intensity distribution pattern but all relevant fragments were observed. 4,5 In the Supplementary Figure   11 the fragmentation of the diastereomeric product mixture presented above is illustrated, indicating that diastereomeric mixtures of molecular keys can be unambiguously read out similar to diastereomeric pure molecular keys.