File Name : fig s1.tif

Caption : fig. s1 liquid chromatogram of a: 5-hmf and b: fructose.

File Name : fig s2.tif

Caption : fig. s2 effect of water content on 5-hmf yield for 1 wt% and 20 wt% fructose concentration. reaction conditions: ■: 1 wt% fructose concentration, ●: 20 wt% fructose concentration, n(hcl)/n(fructose) = 3.6%, 6.1 g 1,4-dioxane, 140 °c, 30 min, rotate speed: 1000 rpm.

File Name : fig s3.tif

Caption : fig. s3 effect of water content on 5-hmf yield in different fructose concentrations. reaction conditions: fructose was used as substrate, n(hcl)/n(fructose) = 3.6%, 6.1 g 1,4-dioxane, 11 wt% h<sub>2</sub>o, 140 °c, 30 min, rotate speed: 1000 rpm

File Name : fig s4.tif

Caption : fig. s4 mass spectrum of unknown intermediate in reaction products a: rt. at 10.85 min and b: rt. at 12.20 min.

File Name : fig s5.tif

Caption : fig. s5 molecular isomers of fructose and their acid catalyzed dehydration to 5-hmf and dfas.

File Name : fig s6.tif

Caption : fig. s6 various pathways of the growth of humins via 5-hmf.

File Name : fig s7.tif

Caption : fig. s7 <sup>13</sup>c-nmr of d-fructose in a: dmso-d6, b: egde:water (8:2), c: dio:water (9:1), d: thf:water (8:2), e: ace:water (8:2), f: d<sub>2</sub>o, g: egme:water (9:1) at 25 °c. note: the addition of water is to make 20 mg of fructose fully dissolved for nmr detection. for thf:water (8:2) mixture, only 10 mg fructose is soluble. we therefore used isotopic 2-<sup>13</sup>c-labeled d-fructose instead to obtained satisfactory nmr signal.

