An Investigation into the Structure and Chemical Properties of Formamidine Sulfinic Acid David Lewis, John Mama, Jamie Hawkes* Perachem Limited, 1 Sizers Court, Henshaw Lane, Yeadon, Leeds, LS19 7DP UK

The use of formamidine sulfinic acid in the textile industry goes back many years, particularly as an agent for the reduction or ‘‘vatting’’ of vat dyes, to form the water-soluble leuco species, when dyeing or printing cellulosic fibers. Many workers have labeled this agent as thiourea dioxide and theorized that its reducing power developed through the existence of two isomeric forms: thiourea dioxide (the sulfone), and formamidine sulfinate. This article has used solid and solution Fourier transform infrared (FT-IR) spectroscopy to show that the name thiourea dioxide is a misnomer since the compound does not exist in the six-valent state, but is stabilized in the four-valent state by formation of a strong internal zwitterion as shown in the article. Index Headings: Formamidine sulfinic acid; Thiourea dioxide; Fourier transform infrared spectroscopy; FT-IR spectroscopy; Formamidinium sulfinate.

INTRODUCTION DeBarry Barnett1 first prepared a compound by the oxidation of thiourea with hydrogen peroxide, which was termed ‘‘thiourea dioxide.’’ Many years later, von Bo¨eseken2,3 carried out a study of the reductive chemical properties of this compound. The use of this product in the textile industry goes back many years particularly as an agent for the reduction or vatting of vat dyes, to form the water-soluble leuco species, when dyeing or printing cellulosic fibers; Lubs4 patented its use for dyeing and printing with vat dyes, discharge printing, and stripping off-shade dyeing. Krug5 described the use of this compound in the printing of vat dyes at pH 5 on cellulose acetate fibers; during steaming the reducing potential is such that the free leuco vat acids are formed; these are not particularly water soluble, but act rather like disperse dyes and hence dye hydrophobic cellulose acetate fibers. This concept was ultimately extended to the dyeing of polyester and nylon fibers with acid leuco vat dispersions.6 The use of ‘‘thiourea dioxide’’ has been favored over the similarly used reducing agent sodium formaldehyde sulfoxylate (sodium hydroxymethane sulfinate) for safety reasons, as the latter produces toxic formaldehyde on use. Krug5 pointed out that the structure of the so-called ‘‘thiourea dioxide’’ was unclear and theorized that its reducing power developed through the existence of two isomeric forms, ‘‘thiourea dioxide’’ (TDO) (the sulfone), and ‘‘formamidine sulfinate’’ (FSA), as shown in Fig. 1. It Received 26 September 2013; accepted 16 May 2014. * Author to whom correspondence should be sent. E-mail: Jamie@ perachem.com. DOI: 10.1366/13-07306

Volume 68, Number 12, 2014

should be noted that the formalized oxidation state of the sulfur atom is þ2 in both compounds, despite the valency change. It was believed that the TDO is converted to FSA only under alkaline conditions, whereupon decomposition would release the reducing sulfinate ion together with urea. Within the textile industry the above explanation was generally accepted as fact and repeated in numerous publications.7–11 However, the unique structure and properties of TDO have also been of academic interest to physical chemists, who found the solid to be unusually densely packed with a very long carbon-sulfur bond. In 1962 Sullivan and Hargreaves12 used X-ray crystallography to conclude that the structure of TDO was in fact the zwitterionic form of FSA, Structure II in Fig. 2. The data presented suggested equal C–N bond lengths with a delocalized positive charge spread across the two N atoms, together with a significant negative charge shared by the oxygen atoms. This conclusion was broadly supported by two further studies in 1990 and 1996 using a combination of X-ray diffraction, molecular mechanics, ab initio, and density functional theory (DFT).13,14 More recently (2003), Denk et al. gave a different interpretation15 and, using geometric and energy parameters derived from DFT, concluded that FSA was the adduct of a carbene with sulfur dioxide (Structure III in Fig. 2) and thus more closely resembles the TDO structure. In 2010, in order to reconcile these conflicting viewpoints, Kis et al. used computational methods to conclude that in the solid state TDO exists as the carbene adduct, and that in solution the more stable tautomer is FSA.16 Much of the evidence presented above is limited in scope and uses only theoretical calculations and X-ray crystallography data. An analytical technique ideally suited to identification of certain functional groups is Fourier transform infrared (FT-IR) spectroscopy. There is a clear structural difference between the structures of TDO and FSA, which centers on whether there is a SO2 group (in the TDO form and the carbene adduct) or a sulfinic acid residue (in the FSA form). Such a distinction is easily identifiable by FT-IR; an FT-IR spectrum of solid FSA is available without comment from many manufacturers, for example, Sigma-Aldrich. The only FT-IR study of this compound dates back to 1965 (Kharitonov and Prokof’eva) in a Russian language journal, which the British Library holds in English in their Cover-to-Cover Translation series.17 This study was concerned with only the solid state FT-IR measurement of FSA. With the benefit of more modern machinery we

0003-7028/14/6812-1327/0 Q 2014 Society for Applied Spectroscopy

APPLIED SPECTROSCOPY

1327

FIG. 2. Postulated structures of formamidine sulfinic acid.

FIG. 1. Historical description of the formamidine sulfinic acid tautomers and the proposed decomposition route to the sulfinate anion.

have replicated the original data and examined the FT-IR spectra of FSA in solution, at various pH values, for the first time. Data from FT-IR analysis presented here suggests that the originally proposed zwitterionic tautomer of FSA is the correct structure, and not the TDO carbene tautomer (Structures II and III, respectively, in Fig. 2). It is interesting to discuss the unusual stability of TDO and FSA. If the sulfinic acid structure is correct, then the citation regarding the low stability of these compounds has to be considered. Oae and Kunieda18 state that ‘‘Sulfinic acids are not very stable and undergo disproportionation or decomposition’’. Marvel et al.19 also state that aliphatic sulfinic acids are much less stable than aromatic sulfinic acids, and the lower alkane sulfinic acids are known to be the least stable. In contrast, TDO and FSA are stable—a property that can be explained by the internal zwitterionic stabilized tautomer of the FSA (Structure II in Fig. 2), and this article summarizes analytical evidence to support this proposal.

EXPERIMENTAL Materials. Formamidine sulfinic acid was supplied by Wego Chemical & Mineral Corp. (USA). All other chemicals were purchased from Sigma-Aldrich (UK) as analytical grade reagents. Methods. Fourier transform infrared spectroscopy was performed on all samples (except those in Fig. 6) using a PerkinElmer FrontierTM Spectrometer with a PerkinElmer UATR attachment. Spectra were handled using the Spectrum 10 software supplied by PerkinElmer (PerkinElmer Instruments, UK). This analysis was kindly performed by PerkinElmer Instruments, UK. The FT-IR analysis of aqueous FSA solutions at variable pH (Fig. 6) was performed using a PerkinElmer Spectrum One Spectrometer with a zinc selenide ATR cell. The water spectra were subtracted using the Spectrum 6 software. Any adjustments in pH prior to measurement of solution spectra were carried out by adding either 30% (w/w) HCl or 50% (w/w) NaOH solutions as appropriate.

RESULTS AND DISCUSSION A central difference between the proposed structures of FSA concerns the bonding of the sulfur atom with the

1328

Volume 68, Number 12, 2014

two oxygen atoms. A sulfinic acid structure has distinct S=O and S–O bonds, or when ionized a delocalized negative charge across these two bonds, whereas the TDO (or carbene adduct) has two S=O bonds and is thus a sulfone-type structure. To see an example of the distinctive sulfone group we have examined methyl phenyl sulfone as a model compound and compared it to the corresponding sulfoxide and sulfide. All three FT-IR spectra can be seen in Fig. 3. All three compounds share medium to weak S–CH3 rocking vibration bands around 1080– 1090 cm1. The sulfide has no strong absorption between 1400 and 1000 cm1, and the sulfoxide shows the single main S=O stretching at 1034 cm1, whereas the sulfone shows the distinctive asymmetrical and symmetrical stretching of an SO2 group at 1282 and 1142 cm1, respectively. These strong bands are typical of all organic sulfones and in the solid phase range from 1360 to 1270 cm1 for the asymmetric stretching vibration and 1170 to 1120 cm1 for the symmetric stretching vibration. Fourier transform infrared data on sulfinic acids and their salts are very hard to obtain and severely lacking in the literature. Data supplied by Socrates,20 however, give the sulfinic acid S=O stretch at 1090–990 cm1 and the S–O stretch at 870–810 cm1, significantly apart from the sulfone absorbance. The sulfinic acid salt is quoted as showing a strong asymmetrical S---O stretch at 1030 cm1 and a strong symmetrical S---O stretch at 980 cm1, the former being the most intense. An examination of the FT-IR spectrum of FSA in solid and in aqueous solution at pH 2.5 (with the water spectrum subtracted) is shown in Fig. 4. Assignment of relevant peaks in solid FSA (Fig. 4a) shows the expected N–H stretching at 3228 and 3017 cm1. Of particular interest is the broad peak at 994 cm1 with shoulder peaks at 1023 and 1040 cm1 due to the sulfinic acid group. This corresponds well to the results from Kharitonov and Prokof’eva,17 who attributed the existence of three rather than two peaks to the influence of the crystalline state on the vibration of the sulfinic acid structure. In the water subtracted aqueous spectrum of FSA (Fig. 4b) the sulfinate structure is resolved into a symmetrical S---O stretch at 1079 cm1 and an asymmetrical S---O stretch at 1000 cm1. In either the solid or solution spectrum of FSA there is no visible sign of any of the characteristic peaks associated with a sulfone-type functional group, namely, in the 1400–1000 cm1 region. As such, it is reasonable to suggest that FSA has a very strong sulfinate character

FIG. 3. FT-IR spectra of methyl phenyl sulfide (a), methyl phenyl sulfoxide (b), and methyl phenyl sulfone (c).

based on the IR spectral signatures and that the carbenoid TDO structure makes little contribution. A comparison of the solid and solution spectra shows very little difference in the state of the sulfinate group. This suggests that the compound exists in a very similar structure in both solid form and aqueous solution,

contradicting Kis et al., who postulated a tautomerism from the carbenoid in the solid state to the sulfinate in solution.16 The remaining bands in the FT-IR spectra belong to the (NH2)2C group. Kharitonov and Prokof’eva found that the absorption at 1682 cm1 in the solid state is relatively

APPLIED SPECTROSCOPY

1329

FIG. 4.

FT-IR spectra of (a) solid FSA and (b) a saturated aqueous solution of FSA (water spectrum subtracted).

unchanged upon deuteration and attributed this peak to the C=N bond of an amidate residue. This is consistent with our results, and the FT-IR spectra of aqueous guanidine hydrochloride provides extra confirmation in that its FT-IR spectra (Fig. 5) show a single amidate peak at 1663 cm1. In FSA we found that this peak shifted slightly to 1701 cm1 when in solution; a similar shift is

FIG. 5.

1330

observed when comparing solid guanidine hydrochloride to an aqueous solution. The broad band in solid FSA at 1432 cm1 is more problematic to characterize. Kharitonov and Prokof’eva17 observed a peak at 1438 cm1 that disappeared on deuteration—a clear indication that this vibration is from the amidate residue and most likely NH2 stretching. It

FT-IR spectrum of aqueous guanidine hydrochloride (water spectrum subtracted).

Volume 68, Number 12, 2014

FIG. 6. Effect of pH on an aqueous solution of formamidine sulfinic acid.

was concluded, despite the crystallographic data at the time indicating the opposite, that the CN bonds in the FSA solid were non-identical with the positive charge localized on a single NH2 group. Since then, the more recent X-ray crystallographic studies back up the original study and maintain that there are identical CN bonds of equal length and therefore a delocalized amidate structure. Our own FT-IR analysis shows the peak at 1432 cm1 in the crystalline form, which in aqueous solution is diminished to a small absorption at 1425 cm1. This evidence is supportive of a non-symmetrical amidate residue with a localized positive charge ‘‘trapped’’ in the solid state and a symmetrical delocalized charge in solution. Again, however, this is in conflict with known crystallographic data, and we have been unable to reconcile the existence of this absorption in the solid state with the X-ray data. Formamidine sulfinic acid is known to be relatively stable in acidic solution; in practice it is only when made alkaline that FSA shows its reducing properties. As stated earlier it was a long-held belief within the textile

industry that the compound exists in the TDO form in acid conditions and tautomerizes to FSA only when made alkaline, where decomposition would rapidly occur. The pH dependence of FSA in aqueous solution was examined by FT-IR, starting at pH 2 and ending at pH 10, as shown in Fig. 6. The remarkable stability of FSA in water between pH 2 and 8 (at 25 8C) is demonstrated by the unchanging FT-IR spectra in Fig. 5. Decomposition does not start to occur until pH increases to 9; at pH 10 there is a complete absence of the amidate group at 1702 cm1 and the SO2 bands are weakened and shifted. These results support the argument that FSA is stabilized by internal zwitterion formation and becomes progressively destabilized with deprotonation of the amidate residue on raising the pH to 9 and above. The pH of decomposition is consistent with known pKa’s of amidate groups. At this pH the strongly reducing HSO2 species appears in solution at 1027 and 960 cm1, together with urea, showing bands at 1630 and 1468 cm1. At no pH value does a sulfone or TDO-type structure appear (note that the doublet at 2360 cm1 is dissolved carbon dioxide).

APPLIED SPECTROSCOPY

1331

It is interesting to note that the SO2 bands are unchanged in the pH 2–8 range indicating that the vibrations are from the sulfinate salt and not the sulfinic acid. Also of note is the remarkable similarity between the SO2 of the sulfoxylate decomposition product generated at pH 10 and the SO2 of the FSA, a further example of their structural similarity.

CONCLUSIONS The data obtained show a compound containing an amidate residue bonded to a sulfinic acid. Organic sulfinic acids are notoriously unstable, and yet FSA in solid state at room temperature is storage-stable for years. Its remarkable stability is due to internal zwitterionic stabilization as demonstrated in Fig. 2 (Structure II); the pKa of the amidate residue will be in the region 9–13, and the pKa of the sulfinic acid is in the region 1–2. Solution (and solid) FT-IR spectra support the presence of the asymmetric/symmetric SO2 bands of a sulfinic acid at 1079 and 1000 cm1, respectively, and not the typical asymmetric/symmetric stretching of an SO2 sulfone group in the proposed TDO/carbene structure. In solution FT-IR spectra, the amidate residue is present (1701 cm1) from pH 2 until the decomposition pH of 9 is reached. The FT-IR evidence presented shows that there is not a proton on the SO2 moiety (under any conditions), and hence the group is a sulfinate. There is very little indication of a carbenoid isomer, and therefore one can conclude that ‘‘thiourea dioxide’’ is not a dioxide or a carbene, but more accurately a sulfinic acid–type structure and thus should be referred to as either ‘‘formamidine sulfinic acid’’ or ‘‘formamidinium sulfinate.’’ ACKNOWLEDGMENTS The authors would like to acknowledge and thank PerkinElmer Instruments (UK) for their kind assistance in performing the FT-IR spectroscopy as described in the Methods section. 1. E.D. Barnett. ‘‘VII.—The Action of Hydrogen Dioxide on Thiocarbamides’’. J. Chem. Soc., Trans. 1910. 97: 63-65.

1332

Volume 68, Number 12, 2014

2. J. von Bo¨eseken. ‘‘The Oxidation Products of Thiourea. The Dioxide as Derivative of Sulphoxylic Acid (H2SO2)’’. Proc. Koninklijke Akad. 1936. 39: 717-721. 3. J. von Bo¨eseken. ‘‘E´tude sur les Oxydes de Thioure´e, I. Sur le Dioxyde de Thioure´e, CS(NH2)2O2’’. Receuil Traveaux Chimique des Pays Bas. 1936. 55(12): 1040-1043. 4. H. August Labs, E.I. du Pont. ‘‘Process for Reducing Vat Dyestuffs’’. US Patent 2164930. Filed 1937. Issued 1939. 5. P. Krug. ‘‘Thiourea Dioxide (Formamidine Sulfinic Acid) a New Reducing Agent for Textile Printing’’. J. Soc. Dyers Color. 1953. 69(13): 606-611. 6. S.N. Chevli, D.M. Lewis. ‘‘The Application of Vat Pigment Dispersions and Leuco Vat Acid Dispersions to Polyester Fibres’’. Adv. Colour Sci. Technol. 1999. 2: 148-151. 7. M. Kwasny. ‘‘Zur Wirkung reducktiver Bleichmittel auf Wolle’’. [Ph.D. Thesis]. Aachen, Germany: Rheinisch-Westfa¨lische Technische Hochschule, 1988. 8. M. Weiss. ‘‘Thiourea Dioxide: The Answer for Safe, Economical and Odourless Reduction in Mill Processing’’. Can. Textile J. 1980. 97: 47-51. 9. M. Weiss. ‘‘Thiourea Dioxide: A Safe Alternative to Hydrosulfite Reduction, Part I’’. Am. Dyest. Rep. 1978. 67(8): 35-38. 10. E.O. Fischer, W. Hieber. ‘‘Zur Kenntnis der Formamidinsulfinsa¨ure’’. Zs. Anorg. Allg. Chem. 1953. 271(3-4): 229-234. 11. P. Krug. ‘‘Thioharnstoffdioxyd (Formamidinsulfinsa¨ure), ein neues Reduktionsmittel fu¨r den Textildruck’’. Textil Rundschau. 1954. 9: 87-88. 12. R.A.L. Sullivan, A. Hargreaves. ‘‘The Crystal and Molecular Structure of Thiourea Dioxide’’. Acta Cryst. 1962. 15: 675-682. 13. Y. Wang, N.-L. Chang, C.-T. Pai. ‘‘Charge Density Study of Thiourea S, S-Dioxide’’. Inorg. Chem. 1990. 29(17): 3256-3259. 14. J.S. Song, E.H. Kim, S.K. Kang, S.S. Yun, I.-H. Suh, S.-S. Choi, S. Lee, W.P. Jensen. ‘‘The Structure and Ab Initio Studies of Thiourea Dioxide’’. B. Korean Chem. Soc. 1996. 17(2): 201-205. 15. M.K. Denk, K. Hatano, A.J. Lough. ‘‘Synthesis and Characterisation of a CarbeneSO2 Adduct—New Insights into the Structure and Bonding of Thiourea S,S-Dioxides’’. Eur. J. Inorg. Chem. 2003. 2: 224-231. 16. Z. Kis, S.V. Makarov, R. Silaghi-Dumitrescu. ‘‘Computational Investigations on the Electronic Structure and Reactivity of Thiourea Dioxide: Sulfoxylate Formation, Tautomerism and Dioxygen Liberation’’. J. Sulfur Chem. 2010. 31(1): 27-39. 17. I. Kharitonov, I.V. Prokof’eva. ‘‘Infrared Absorption Spectra and Structure of Thiourea Dioxide’’. Dokl. Akad. Nauk. SSSR. 1965. 162(4): 829-832. 18. S. Oae, N. Kunieda. ‘‘Sulphinic Acids and Sulphinic Esters’’. In: S. Oae, editor. Organic Chemistry of Sulphur. New York: Plenum Press, 1977. Pp. 603-647. 19. C.S. Marvel, R.S. Johnson. ‘‘1-Dodecanesufinic Acid’’. J. Org. Chem. 1948. 13(6): 822-829. 20. G. Socrates. Infrared and Raman Characteristic Group Frequencies: Tables and Charts. Chichester, UK: John Wiley and Sons, 2004. 3rd ed.

An investigation into the structure and chemical properties of formamidine sulfinic acid.

The use of formamidine sulfinic acid in the textile industry goes back many years, particularly as an agent for the reduction or "vatting" of vat dyes...
329KB Sizes 0 Downloads 6 Views