Synthesis and tastant properties of disulfamates. Multisulfamation studies

Abstract

Twenty four disulfamates, one trisulfamate, two tetrasulfamates and two monosulfamates have been made. The disulfamates are of two types: RN(SO₃Na)₂ (Type A, compounds 1–20) and NaO₃S(H)NR′N(H)SO₃Na (Type B, compounds 21–24) and all except three (which had not been tasted) are new materials. The positions of the –SO₃Na groups in compounds 21–23 have been established by the use of model compounds (e.g. parent amines, appropriate monosulfamates) and ¹³C-NMR. Some multisulfamation synthesis leading to compounds 25–27 has been carried out. Taste data have been obtained for almost all the sulfamates made and the significance of these in relation to structure–taste studies for sulfamate sweeteners is discussed. In particular, the possibility that the entity [double bond splayed left] CHN(R)SO₃⁻ might function as a hydrogen source in the Shallenberger–Acree, multicomponent attachment and α-helical protein receptor mechanisms has been examined.

---

There are over 30 reports in the literature dealing with disulfamates (imidodisulfonates) ranging from the initial work in 1920 of Traube and co-workers ^1,2 [who made some basic compounds such as RN(SO₃K)₂ (R = Me, Et, Pr) ^1,2 and simple hydroxylamine disulfonates such as MeON(SO₃K)₂ ^2,3] to a current report ⁴ where 2,2-dinitropropane-1,3-diol [HOCH₂C(NO₂)₂CH₂OH] reacts with alkylsulfamates (RNHSO₃M) to give, inter alia, the disulfamates MO₃SN(R)CH₂C(NO₂)₂CH₂N(R)SO₃M. Most reports deal with ionic disulfamates but there are also some reports ^5–8 of various disulfamate esters and one ⁹ of a ‘mixed’ disulfamate e.g. a monobactam which contains a sulfamate ester group (see Fig. 1).

Fig. 1 Some previously reported disulfamate esters (refs. 5–9).

There are a number of reports dealing with tetra- and trisulfamates.^1,10–13 Traube and Wolf f ¹ report the preparation of the tetrasulfamate (KO₃S)₂NCH₂CH₂N(SO₃K)₂. Various hydrazine tri- and tetrasulfonates have been synthesized ¹¹ and in two papers ^12,13 Baumgarten describes the preparation of some tetra potassium and barium sulfonate salts of urea,¹² a series of aminoacid disulfonates and one trisulfonate, N,N ′,N ′′-histidylhistidine trisulfonate.¹³

Few reports have focused on the syntheses of disulfamates per se apart from the work of Baumgarten,¹³ who synthesized seven di-/tri- amino acid sulfonates using pyridine–sulfur trioxide, and Kanetani,^14,15 who used amine–SO₃ adducts and amines to synthesize over 20 disulfamates including 1–3 (Fig. 2). Kanetani carried out a detailed study of the effect of the reaction time, temperature etc. on the disulfamation reaction.

Fig. 2 Type A disulfamates synthesized; Z = SO₃Na.

Disulfamates have found application in a number of areas. Thus, alkanediyl disulfamates [H₂NSO₂O(CH₂)_nOSO₂NH₂, n = 6, 7, 8] inhibit the growth of malignant cells in mammals,¹⁶ the monobactam sulfamates shown in Fig. 1 are reported to have in vitro antibacterial activity,⁹ various derivatives of 3,5-bis(sulfoamino)benzoic acid act as anti-inflammatory agents in the treatment of rheumatic fever and rheumatoid arthritis ¹⁷ and a number of disulfamates are intermediates in the dyeing industry.¹⁸ Finally the disulfamate aniline-N,N-disulfonic acid, PhN(SO₃H)₂ has been implicated in the mechanisms of both the sulfonation of aniline in conc. sulfuric acid and the rearrangement of phenylsulfamic acid ¹⁹ and in the reduction of nitro compounds with sulfite (Piria reaction).²⁰

Disulfamates could, since they contain a sulfamate moiety, be sweet compounds and this indicates another possible use. Unfortunately, there are no taste data at all available for all the disulfamates already prepared. In this paper we explore this potential by synthesizing twenty one new disulfamates and three previously made by Kanetani ¹⁴ (see Figs. 2 and 3) together with one trisulfamate, and two tetrasulfamates (see Fig. 3). The taste portfolios of 25 of the compounds together with those of two monosulfamates also made in this work for comparison have been assessed.

Fig. 3 Type B disulfamates (21–24) and multisulfamates (25–27) synthesized; Z = SO₃Na.

Results and discussion

The sweet taste elicited by a number of metallic sulfamates [RN(H)SO₃⁻M⁺] has been known for about 60 years.²¹ Structure–taste studies have been carried out on sulfamates synthesized with alicyclic, open-chain, aromatic, heterocyclic and hetero open-chain and branched R groups, and a number of the compounds were found to be sweet.²² The cation, M⁺ does not appear to effect the degree of sweetness of a sulfamate.²² An intact –NHSO₃⁻ function was considered to be essential for sweetness since it provides a hydrogen and an electronegative center for the operation of the Shallenberger–Acree mechanism of sweetness.²³ However, some years ago a series of heterocyclic sulfides and sulfones containing an –N(R)SO₃⁻ group were reported to be sweet,²⁴ and their sweetness has been explained on the basis of the presence of the system [double bond splayed left]CHN(R)SO₃⁻, in which an α-hydrogen might function as a hydrogen source in the Shallenberger–Acree mechanism.²⁵ The reported sweetness of these compounds has more recently been disputed.²⁶

We were interested in testing the idea that an α-hydrogen, rather than a hydrogen on nitrogen, might be able to participate in the sweetness mechanism. The synthesis of disulfamates seems to offer an ideal opportunity to test this since (i) the amino hydrogen is replaced, (ii) a second sulfamate entity is being introduced into the molecule and, given the well-established “sweet-conferring ability” of the sulfamate group, it might enhance sweetness and (iii) we have been able to show that one of the α-hydrogens in CH₃N(SO₃K)₂ is within a distance of 3 Å of two of the oxygen atoms and it could therefore act as hydrogen source in the operation of the Shallenberger–Acree mechanism. This was established by downloading from the Cambridge Structural Database the structure for the disulfamate K₂[CH₃N(SO₃)₂] (no. 39971) (see Fig. 4) determined by Kennard et al.²⁷ Using the ORTEX (OSCAIL) program ²⁸ we obtained the relevant interatomic distances and found that H11–O12 was 2.665 and H11–O11 was 2.916 Å. All other H–O distances were >3 Å and were not considered. It should be noted that H13 was not located in the crystallographic work of Kennard.

Fig. 4 Schematic representation of K₂[CH₃N(SO₃)₂] from the Cambridge Structural Database.

Two types of disulfamate have been prepared—Type A [RN(SO₃Na)₂, i.e. both sulfonates on one nitrogen, N,N-disulfamates] and Type B [NaO₃S(H)NR′N(H)SO₃Na, i.e. the two sulfonates on different nitrogens, N,N ′-disulfamates].

All the disulfamates prepared (Figs. 2 and 3) have one or two α-hydrogens at C1, except compound 20 (Fig. 2), which was included because it had not been prepared by Kanetani when he prepared a series of aryldisulfamates.¹⁵ Of the 29 compounds prepared in this work only three compounds (1–3) have been previously prepared.^1,14 These were resynthesized as standards and also because their taste data was needed for comparison with their monosulfamates, two of which are sweet.

Taste portfolios for 25 compounds, together with percentage yields (for 27 compounds) are given in Table 1. From the taste data in Table 1 it is readily apparent that, though some of the compounds are multisapophoric, none of the 25 compounds tasted display sweetness, apart from a sweet aftertaste in compound 22. The pHs of most of the 0.01 M solutions for tasting were determined and these are given in Table 1. As would be expected,²³ those solutions that have a predominant sour taste display a definite acidic pH, while bitter solutions give a pH in the alkaline region. The solutions found to be tasteless, viz. those of compounds 14, 19 and 23, gave pHs of 10.2, 10.3 and 5.9, respectively. In the tasting of compound 5 sourness seemed to predominate over bitterness, thus the pH of 2.3 is reasonable. The one peculiarity arises with the naphthyldisulfamate (20) which, despite being exclusively bitter, displays a solution pH of 3.00. One notes, however, that a secondary taste of mothballs was found with this compound and this may be masking a sour taste. In Table 2 data (where available) for the corresponding monosulfamates are compared with those for the disulfamates. In the case of compounds 22 and 23 taste data for the corresponding monosulfamates (28 and 29) were obtained in this work (see footnotes, Table 2). From Table 2 it is seen that for compounds 2, 3 and 23 there is a change in taste on going from the mono- to the disulfamates and for 22 there is no change. Comparison is difficult for other compounds, as when most of the monosulfamate taste data was collected between 11 and 37 years ago it was common to describe compounds that did not give a sweet taste as simply nonsweet,^25,28,34 non sucre,^29,33 or non dolce.³⁵ However, for compounds 1, 4, 9, 12, 13, 15, 17–19 and 22 sweetness is absent in the monosulfamates and disulfamation does not induce sweetness. Simple meta-substituted phenylsulfamates are often sweet and hence some sweetness might be expected in 21, but not in 22, as sweetness has not been found in para- or ortho-phenylsulfamates. Some sweet taste might also be expected in 23 and in its monosulfamate 29 since some monosubstituted cyclamate derivatives are sweet,³⁶ and in fact the monosulfamate 29 has a substantial sweet component.

Table 1 Percentage yield, pH, taste data ^a and percentage of tasters giving the taste quality of disulfamates 1–27

Compound	Yield (%)	pH	Sour	Bitter	Salty	Tasteless	Aftertaste (% of tasters)	No. of tasters	Predominant taste (>50% tasters)
a All compounds were tasted as 0.01 M aqueous solutions and allowance was made for water of hydration when making up solutions. pH measurements were made using these solutions and an Orion pH meter model 420A buffered at pH 4.0, 7.0 and 9.2. Taste panellists did not detect an initial sweet taste in any of the solutions. b An onion-like taste was also reported for this compound. c A musty taste was also reported. d A taste of mothballs was also reported. e A taste of nuts was also reported. f Not tasted. g Taste of burnt nuts was also reported.
1	42	2.35	100	0	0	0	0	9	Sour
2	15	—	100	0	0	0	0	4	Sour
3	12	—	100	0	0	0	0	4	Sour
4	45	6.2	45	0	67	0	0	9	Salty
5	41	2.3	67	55	0	0	0	9	Sour/bitter ^b
6	20	2.2	89	45	0	0	0	9	Sour
7	21	2.2	100	0	0	0	0	9	Sour
8	60	10.4	0	0	100	0	0	9	Salty
9	16	11.1	0	100	0	0	0	9	Bitter
10	25	11.3	22	78	0	0	0	9	Bitter
11	20	10.8	0	100	0	0	Bitter (33)	9	Bitter
12	45	—	100	0	0	0	0	4	Sour
13	20	8.1	78	22	0	0	0	9	Sour
14	68	10.2	45	0	0	55	0	9	Tasteless
15	15	2.5	78	0	0	22	0	9	Sour
16	54	10.5	0	55	0	45	0	9	Bitter ^c
17	33	—	0	100	0	0	0	4	Bitter
18	36	10.3	0	100	0	0	0	9	Bitter
19	29	10.3	0	0	0	100	0	9	Tasteless
20	30	3.00	0	100	0	0	0	9	Bitter ^d
21	34	—	0	100	0	0	0	4	Bitter
22	46	—	25	100	0	0	Sweet (25)	4	Bitter
23	15	5.9	0	0	0	100	0	9	Tasteless ^e
24	10	—	—	—	—	—	—	—	— ^f
25	13	—	100	0	0	0	0	9	Sour
26	7	—	0	100	0	0	0	9	Bitter ^g
27	10	—	—	—	—	—	—	—	— ^f

---

Table 2 Comparison of predominant tastes of di- and multisulfamates with the corresponding monosulfamates

	Taste
Compound	Di-	Mono-	Ref.^c
a Panellists (4) found this compound to be 100% bitter and to have 25% sweet aftertaste only—therefore its predominant taste is bitter. b Panellists (4) found this compound to be 100% bitter and 100% sweet—therefore its predominant taste is recorded as bitter/sweet. c Reference to monosulfamate.
1	Sour	Nonsweet	25
2	Sour	Sweet	25,29,30
3	Sour	Sweet	25,29–31
4	Salty	Nonsweet	29
9	Bitter	Bitter	32
12	Sour	Nonsweet	33
13	Sour	Nonsweet	34
15	Sour	Nonsweet	33
17	Bitter	Nonsweet	35
18	Bitter	Nonsweet	35
19	Tasteless	Nonsweet	32
22 (di), 28 (mono)	Bitter	Bitter ^a	Present work
23 (di), 29 (mono)	Bitter	Bitter/Sweet ^b	Present work

---

The di-, tri- and tetrasulfamates 24, 25 and 26 and 27, respectively, are long-chain compounds and would be unlikely to be sweet since even a C5 chain hydrocarbon sulfamate is not sweet.²⁹

The absence of sweetness in the disulfamates does appear to underline the necessity for an imino hydrogen (–NH) in the sulfamate functionality that can actively participate in hydrogen bonding in the dual H-bond theory of Shallenberger and Acree. Substitution of an α-hydrogen, as we have tired to do with the disulfamates, does not work, even though we have shown above that the distance between the –CH– and the oxygen can be within 3 Å, the required AH–B distance of the Shallenberger–Acree theory. A C-bound hydrogen would, of course, be much less efficient in H-bonding than an –NH or –OH and this is probably an important factor too. In terms of the multicomponent attachment theory of sweetness of Nofre and Tinti,³⁷ one might assign the alkyl or alkoxy or piperidine or morpholine rings (Fig. 2) as G sites, the α-CH as an AH site and –SO₃– as a B site. On this basis the AH–G distance could be ≈3.5 Å and the B–G distance ≈5.5 Å, respectively, as required by the theory. One selects a suitable point in G to achieve this. It is difficult to assign other sites (the theory normally requires at least 8 sites) unless one takes one of the SO₃⁻ groups as a D site.

Using their α-helical protein receptor theory, Suami and Hough ³⁸ have recently examined some cyclamate derivatives and explained very well the sweetness of some and the lack of sweetness in others. In the disulfamates it is relatively easy to pick a glycophoric triad of AH_s/B_s/X_s (using their notation). Thus, AH_s would be the α-CH, B_s one of the –SO₃⁻ groups and X_s appropriate hydrogens in the R portion. In so far as one can see, therefore, the three main sweetness theories would indicate that the disulfamates might elicit sweetness. The failure to realize this may well be due to the weak H-bonding ability of the α-H_s.³⁹

In summary, 29 sulfamates, mainly disulfamates, have been prepared in this work and taste profiles have been obtained for 27 of these. No sweet taste was noted in the disulfamates 1–19, all of which contain the entity [double bond splayed left]CHN(SO₃Na)₂ (in which the α-hydrogen attached to carbon might be a source of hydrogen, in place of a hydrogen on nitrogen, for the successful operation of the Shallenberger–Acree multicomponent attachment or α-helical receptor theories of sweetness). Further, the sweetness of several monosulfamates is shown to be destroyed in the corresponding disulfamates and a sweet taste could not be “induced” in a number of nonsweet monosulfamates on their conversion to the corresponding disulfamates. In conclusion, this work is in line with the main body of evidence from earlier studies ^21,25,40,41 that indicates that an amino hydrogen is essential for sweetness and it appears to cast additional doubt on the 1974 report that some secondary sulfamates containing an –N(R)SO₃ group were sweet.²⁴

Experimental

Materials and methods

General. All amines used were commercially available. Liquids were distilled using normal distillation apparatus at atmospheric pressure or using a Kugelrohr distillation unit if the amine had a high boiling point and/or where only small quantities were available. Solid amines were recrystallized using appropriate solvents and dried over P₂O₅ for a period of at least 24 h. 2-Picoline was dried over KOH pellets for 24 h, refluxed, distilled and again stored over KOH pellets. Chlorosulfonic acid was distilled at atmospheric pressure and stored in a dark environment in a suitably-sized bottle. 1-(2-Aminoethyl)piperidine was a gift from Allied Colloids Ltd.

Preparation of disulfamates. The synthesis of the disulfamates was based on Kanetani’s procedures.^14,15 Typically, chlorosulfonic acid (0.1 mol, 6.7 ml) was added dropwise to stirred 2-picoline (1.0 mol, 81 ml) at 0 °C under anhydrous conditions. The solution was maintained at 0 °C using an ice-bath with the aid of salt–acetone. The 2-picoline–sulfur trioxide complex was allowed to warm to room temperature with stirring and the amine (0.05 mol) in the minimum amount of 2-picoline was added slowly. The reaction was stirred overnight, after which it was basified to pH ≈ 10 using 3 M NaOH. The resulting solution was washed with diethyl ether several times to extract unreacted amine. The remaining aqueous solution was evaporated in vacuo to yield the crude disulfamate product. This crude product could then be purified by successive recrystallizations with aqueous ethanol with a pH > 7 (lower pH values tend to cleave off one of the sulfamate groups to give monosulfamate). Recrystallization was repeated until the product was free of sulfate and chloride and gave a clean sulfamate test.²⁵ The purified material was dried over P₂O₅in vacuo for at least 24 h.

The 2∶1 ratio of chlorosulfonic acid to amine was stepped up to 2.5∶1 or 3∶1 in order to improve yields for the synthesis of compounds 4, 9, 11 and 18–20. If more than 10 ml of chlorosulfonic acid was used a MeOH–dry-ice bath was used to lower and maintain the temperature and to allow more rapid addition of the acid.

Preparation of other sulfamates. The monosulfamates 28 (9.2% yield) and 29 (11.3%) were synthesized using a 1∶1 ratio of chlorosulfonic acid to amine. Interestingly the same ratio easily trisulfamated N′-methylethylenediamine to give 25. Tetrasulfamates 26 and 27 were obtained by using a 4∶1 ratio of chlorosulfonic acid to amine.

Elemental analyses. Most of the compounds prepared tended to recrystallize with some water of crystallization. This is commonly the case with sulfamates despite lengthy drying.^21,25,29 Microanalyses (C, H and N) for all 29 sulfamates was carried out and S analysis was also performed on compounds 2, 3, 12, 17, 21, 22, 24, 28 and 29. C, H, N and S were within the normal limits (±0.5%) except for the following: compound 5, % N found was +0.59; compound 12, % N found was −0.69 and % S found was −0.64; compound 21, % C found was +0.77; compound 22, % C found was +0.60 and % S found was −0.90; compound 24, % S was −0.73; and compound 27, % N found was −0.58. These six compounds gave good IR and NMR spectra, performed well in the sulfamate test and were free of sulfate and chloride ions, and so were used in tasting.

Spectroscopic analyses. IR spectra of all the sulfamates prepared were recorded and these showed the usual SO₃ and NS bands and those sulfamates containing –NH(s) gave absorbances in the range 3400–3190 cm⁻¹. Spectra were recorded on a Perkin-Elmer 983G spectrometer. Proton and carbon-13 NMR were recorded on JEOL 270 and 400 MHz spectrometers, in ppm relative to TMS. Compounds 1–20 and 25–27 did not give an amino H signal in their ¹H NMR spectra but compounds 21–24 did show such a signal. Compounds 21–24 may form disulfamates of Type B, having –SO₃Na groups on different nitrogens as illustrated for compounds 21–23 (N,N′-disulfamate), or they could have the two –SO₃Na groups on the same nitrogen (N,N-disulfamate) as shown for compound 24 in Fig. 3. For 21δ_C [(CD₃)₂SO] 105.56, 108.69, 128.11 and 143.64 and δ_C for the starting amine phenylene-1,3-diamine [(CD₃)₂SO] 100.18, 103.28, 129.28 and 149.09. Both of these spectra exhibit four carbon signals characteristic of a symmetrically substituted 1,3-disubstituted phenyl ring. Had disulfamation taken place on the same nitrogen, a ¹³C spectrum with six signals would be anticipated. The signals furthest downfield in both spectra are due to the two tertiary C–N carbon atoms and the upfield shift of 5.45 on sulfamation is due to shielding.

In Table 3¹³C chemical shifts are given for N-phenylphenylene-1,4-diamine, the monosulfamate, 28, and the disulfamate product, 22. For the diamine spectrum the assignments have been made from calculation of theoretical ¹³C chemical shifts for this compound using the empirical rules and tables of data of Williams and Fleming.⁴² The assignments of the signals at 132.0, 144.0 and 146.9 to the three tertiary carbons were confirmed by the results of a C-J resolved spectrum. For the monosulfamate the assignments are also shown and again a C-J resolved spectrum confirmed that the three downfield peaks are due to the three tertiary carbon atoms. That the monosulfamate 22 is as shown above is evident by the fact that the C1 signal is, as expected, shifted upfield by 5 ppm, C7 is virtually unaffected and C4 shows an upfield shift of 3.4. Had sulfamation occurred at the NH between the two rings, considerable shifts in the C7 and C4 signals would have been expected. For 21 a C-J resolved spectrum again assigned the three downfield signals as shown. The disulfamate shows a strong upfield shift (6.2 ppm) at C4, whereas at C1 it is only 3.7 and this indicates that it is the N,N′- and not the N,N-disulfamate that has been formed.

Table 3 ¹³C Chemical shifts ^a for N-phenyl-1,4-phenylenediamine, the monosulfamate 28 and the disulfamate 22 relative to Me₄Si

(CD3)2SO	C8/C12	C3/C5	C2/C6	C10	C9/C11	C4	C1	C7
a 270 MHz.
4-Ph(H)NC6H4NH2	114.0	115.1	117.3	122.7	129.2	132.0	144.0	146.9
28	115.1	118.4	118.7	121.1	130.0	135.4	139.0	146.7
22	116.8	117.3	120.3	123.1	123.6	129.2	142.7	143.1

---

For 23δ_C [D₂O, 100 °C] 31.72 and 57.4; for 1,4-diaminocyclohexane δ_C [(CD₃)₂SO] 35.26 and 50.0 and for the monosulfamate, 29, δ_C [(CD₃)₂SO] 32.6, 35.1, 49.9 and 52.1. The spectrum of compound 23 had to be run in D₂O to increase its solubility because it was insoluble in (CD₃)₂SO. Clearly again, as in the case of 21, 23 is an N,N′-disulfamate, and not an N,N-disulfamate, which would give four signals in its ¹³C NMR. Compound 24 gave 10 signals in the ¹³C NMR so it clearly has to be the N,N-disulfamate as shown in Fig. 3.

Acknowledgements

The authors thank Mr N. Ní Chonchubhair for doing the ORTEX calculations. C. A. R. and L. M. K. thank Enterprise Ireland for grants. The EU is thanked for a grant.

References

1. W. Traube and M. Wolff, Ber. Dtsch. Ges. Chem., 1920, 53, 1493.
2. W. Traube, H. Ohlendorf and H. Zander, Ber. Dtsch. Ges. Chem., 1920, 53, 1477.
3. It should be noted that salts of the parent compound imidodisulfonic acid (aquo-ammonosulfuric acid) have been known since 1834. See L. F. Audrieth, M. Sveda, H. H. Sisler and M. J. Butler, Chem. Rev., 1940, 26, 49 for a review..
4. V. A. Tartakovsky, A. S. Ermakov and O. N. Varfolomeeva, Russ. Chem. Bull., 1999, 48, 1385.
5. M. Goehring and H. K. A. Zahn, Chem. Ber., 1956, 89, 179.
6. M. Hedayatullah and J. F. Brault, Phosphorus Sulfur, 1981, 11, 303.
7. N. S. Zefirov, N. V. Zyk and A. G. Kutateladze and Yu. A. Lapin, Zh. Org. Khim., 1987, 23, 229.
8. M. A. Pudovik, L. K. Kibardina and A. N. Pudovik, Russ. J. Gen. Chem., 1993, 63, 267.
9. G. Guanti, R. Riva, G. Cascio, E. Manghisi, G. Morandotti, G. Satta and R. Sperning, Farmaco, 1998, 53, 173; G. Guanti, R. Riva, G. Cascio, E. Manghisi, G. Morandotti, G. Satta and R. Sperning, Chem. Abstr., 1998, 129, 1754465.
10. Salts of the nitrilosulfonates, N(SO₃H)₃ have been known since the last century. See Audrieth et al., reference 3..
11. A. Meuwsen and H. Tischer, Z. Anorg. Allg. Chem., 1958, 294, 282.
12. P. Baumgarten and I. Marggraff, Ber. Dtsch. Ges. Chem., 1931, 64, 301.
13. P. Baumgarten, I. Marggraf and E. Dammann, Z. Physiol. Chem., 1932, 209, 145.
14. F. Kanetani and H. Yamaguchi, Bull. Chem. Soc. Jpn., 1974, 47, 2713.
15. F. Kanetani, Bull. Chem. Soc. Jpn., 1986, 59, 952.
16. D. M. Bailey, Sterling Drug Co., U.S. Patent 4 829 084, 1989; ; D. M. Bailey, Chem. Abstr., 1989, 111, 120951..
17. UCLAF, Br. Patent 879 050, 1961;; Chem. Abstr., 1962, 56, 14173b..
18. (a) R. Lantz, Bull. Soc. Chim. Fr., 1948, 489; (b) F. Ackerman, U.S. Patent 2 567 796, 1951; F. Ackerman, Chem. Abstr., 1952, 46, 754; (c) R. Lantz and P. M. J. Obellianne, Br. Patent 753 983, 1956; R. Lantz and P. M. J. Obellianne, Chem. Abstr., 1957, 51, 5826e; (d) R. F. M. Sureau and P. M. J. Obellianne, U.S. Patent 2 853 359, 1958; R. F. M. Sureau and P. M. J. Obellianne, Chem. Abstr., 1959, 53, 3721h..
19. Z. Vrba and Z. J. Allan, Tetrahedron Lett., 1968, 4507.
20. Z. Vbra and Z. J. Allan, Collect. Czech. Chem. Commun., 1968, 33, 431.
21. L. F. Audrieth and M. Sveda, J. Org. Chem., 1944, 9, 89.
22. W. J. Spillane, C. A. Ryder, M. R. Walsh, P. J. Curran, D. G. Concagh and S. N. Wall, Food Chem., 1996, 56, 255.
23. R. S. Shallenberger, Taste Chemistry, Blackie, Academic & Professional, New York, 1993..
24. G. R. Wendt and M. W. Winkley, U.S. Patent 3 787 442, 1974;; G. R. Wendt and M. W. Winkley, Chem. Abstr., 1974, 80, 82669.
25. G. A. Benson and W. J. Spillane, J. Med. Chem., 1976, 19, 869.
26. See W. J. Spillane, C. A. Ryder, P. J. Curran, S. N. Wall, L. M. Kelly, B. G. Feeney and J. Newell, J. Chem. Soc., Perkin Trans. 2, 2000, 1369 for a discussion of this matter..
27. A. J. Morris, C. H. L. Kennard and J. R. Hall, Inorg. Chim. Acta, 1982, 62, 247.
28. P. McArdle, J. Appl. Crystallogr., 1995, 28, 65.
29. C. Nofre and F. Pautet, Bull. Soc. Chim. Fr., 1975, 686.
30. H. Yamaguchi, Nippo Chemical Co., Jpn. Patent 13,056, 1963;; H. Yamaguchi, Chem. Abstr., 1964, 60, 2771a..
31. B. Unterhalt and L. Böschemeyer, Z. Lebensm. Unters. Forsch., 1972, 149, 227.
32. W. J. Spillane, G. McGlinchey, I. Ó Muircheartaigh and G. A. Benson, J. Pharm. Sci., 1983, 72, 852.
33. F. Pautet and M. Daudon, Pharmazie, 1985, 40, 428.
34. W. J. Spillane and M. B. Sheahan, J. Chem. Soc., Perkin Trans. 2, 1989, 741.
35. M. DeNardo, C. Ruuti and F. Ulian, Farmaco, Ed. Sci., 1984, 39, 125; M. DeNardo, C. Ruuti and F. Ulian, Chem. Abstr., 1984, 100, 137568.
36. W. J. Spillane and G. McGlinchey, J. Pharm. Sci., 1981, 70, 933.
37. J. M. Tinti and C. Nofre, in Sweeteners Molecular Design and Chemoreception, ed. D. E. Walters, F. T. Orthoefer and G. E. DuBois, ACS Symposium Series 450, American Chemical Society, Washington, DC, 1991, ch. 15..
38. T. Suami, L. Hough, T. Machinami, T. Saito and K. Nakamura, Food Chem., 1998, 63, 391.
39. G. A. Jeffrey and W. Saenger, Hydrogen Bonding in Biological Structures, Springer Verlag, 1994, Study Edition, ch. 10..
40. F. F. Blicke, H. E. Millson and N. J. Doorenbos, J. Am. Chem. Soc., 1954, 76, 2498.
41. C. D. Hurd and N. Kharasch, J. Am. Chem. Soc., 1946, 68, 653.
42. D. H. Williams and I. Fleming, Spectroscopic Methods in Organic Chemistry, McGraw-Hill, London, 4th edn., 1987, pp. 125–149..

---