A new application of the traditional Fenton process to gold cyanide synthesis using acetonitrile as a cyanide source

Abstract

Previous cyanation processes always involved the use of hypertoxic CN⁻ ions or UV-light, which is health-risky and energy-wasting. For the purpose of environment-friendly chemistry, we introduced a general Fenton-reaction synthesis of metal cyanide solids using acetonitrile as a green cyanide source. This Fenton-improved cyanation method gets rid of CN⁻ ions and UV-light efficiently, which is green and facile.

---

Metal cyanides (MCs) have a variety of applications in many industrially important processes, such as metallurgy, electroplating, plastics, medicine, pesticides, dyeing, and rubber manufacturing.^1–5 The conventional preparation of MCs,^6–11 however, always utilize highly toxic NaCN, KCN or HCN as cyanide sources,¹² which could potentially threaten humans and the environment even when handled with extreme caution. For the purpose of green chemistry, efforts have been made to find alternative hypotoxic cyanide sources and green ways to produce MCs. Radical-involved cyaniding processes thus came into being.^13,14 The radical-involved processes replace CN⁻ with ˙CN, which not only exclude the toxicity from free cyanide anions, but greatly enrich the cyanide sources, such as acetonitrile (MeCN).

MeCN is conventionally treated as an inert cyanating agent because of its stable C–CN bond (133 kcal mol⁻¹) and has been widely used as polar aprotic solvent in organic synthesis and purification.^15,16 However, our recently publications showed that a photo-induced hydroxyl radical (˙OH) is able to activate the C–CN bond of MeCN to make it a good cyanide source for the production of MCs.¹³ Combining electron paramagnetic resonance (EPR) experiments with DFT calculations, we find that the hydroxyl radical (˙OH) is likely to be the initiator of this photo-assisted cyanation process. It could initiate the reaction in two reasonable pathways. (I) ˙OH would abstract one H atom from MeCN to yield a ˙CH₂CN intermediate, and ˙CN would be produced during the chain propagation process. AuCN would be formed through the combination of ˙CN radicals and AuNPs. (II) ˙OH would attack AuNPs directly to form AuOH intermediates, which would react with nitrile to yield AuCN. ˙OH is indispensable in both reaction routes. To rule out the possible effect of UV irradiation and further identify the role of ˙OH, a UV-absent process to produce ˙OH is desired, which drew our attention to Fenton’s reagent.

Fenton’s reagent, an aqueous mixture of Fe²⁺ and H₂O₂, is believed to function as an oxidizing agent through the formation of ˙OH (Fe²⁺ + H₂O₂ → Fe³⁺ + ˙OH + OH⁻).^17–19 The strong oxidizing ability of ˙OH makes it a powerful agent in the treatment of toxic organic pollutants including benzene, nitrobenzene and phenols.^20–23 While efforts have been made to improve the performance of Fenton’s reagent for the degradation of organic pollutants, applications of Fenton’s reagent in other aspects are rare, especially in non-aqueous systems.

In this paper, Fenton’s reagent is adopted as an ˙OH source to conduct the cyanation of metal nanoparticles. The controllable production of ˙OH as well as the simple and mild reaction conditions make it a perfect tool to investigate the role of ˙OH in the cyanation process. The experimental results indicate that this Fenton-improved cyanation (FIC) method could effectively exclude the toxicity from free cyanide anions. Compared to our previously reported photo-assisted cyanation process, the FIC process avoids the use of UV-irradiation and is thus more energy-saving and more feasible for the large-scale industrial production of MCs. Unlike traditional Fenton processes, which are usually conducted in acidic aqueous solution, the FIC process takes place in MeCN solution, which provides a possibility for Fenton’s reagent to be used in organic synthesis.

In our previous publication, ˙OH radicals were generated in solution through UV-assisted autoxidation of benzaldehyde. The complex mechanism, uncontrollable radical composition and concentration limited our basic understanding of the intrinsic activity of ˙OH radicals in the cyanation of AuNPs, although EPR experiments and DFT calculations provided us with a qualitative description. Benefiting from the well-investigated Fenton process,^24–26 herein, the concentration of ˙OH radical is tuned by adjusting the dosage of Fe²⁺. Typically, ˙OH radicals were produced by mixing 200 μl of 30% H₂O₂ with X mg of FeSO₄·7H₂O (X = 0, 1, 2, 5, 10, 50) in 4 mL of MeCN under air under stirring at 30 °C. 60 mg of 5.0 wt% AuNPs/SiO₂ was added into the solution as the reaction substrate. After reacting for several hours, the solid product was obtained by centrifuging and washing twice with deionized water before drying in a vacuum oven at 60 °C overnight and X-ray diffraction was employed to determine the conversion of Au to AuCN. The control experiments suggest that the conversion could take place only in the presence of both FeSO₄·7H₂O and H₂O₂, implying the indispensable role of the ˙OH radical. Besides, the complete cyanation of 60 mg of 5.0 wt% AuNPs/SiO₂ could be achieved within 4 h only if the addition of FeSO₄·7H₂O was above 10 mg. The diffraction peaks at 2θ = 17.4°, 30.4°, 35.3°, 47.3°, 53.4°, 57.2° and 63.2° could be well indexed to the (001), (100), (002), (102), (110), (111) and (200) plane reflections of a hexagonal AuCN crystal (JCPDS card number: 11-0307). When the addition of FeSO₄·7H₂O is below 10 mg, visible peaks at 2θ = 38.1°, 44.4° and 64.6°, corresponding to (111), (200), and (220) of Au (JCPDS: 04-0784) are detected, suggesting the presence of unreacted Au. The gradually disappearing Au peaks shown in Fig. 1a, along with the decreased time for the complete conversion of AuNPs to AuCN (Fig. 1b), indicating that the cyanation reaction rate is determined by the amount of FeSO₄·7H₂O, i.e. the concentration of ˙OH. It is surprising to find that no other by-products, such as Fe₇(CN)₁₈ (Prussian blue), have been observed in the FIC reaction system despite the presence of a large amount of iron salts (Fig. S1†). This observation indirectly excludes the existence of toxic CN⁻ anions in our reaction system and argues for the radical nature of the FIC process. Besides, only CO₂ and O₂ were detected via GC-TCD in the gas-phase of the FIC system, which further confirmed the green nature of the FIC method.

Fig. 1 (a) XRD patterns of the solid products obtained using the FIC method after a 4 h reaction with different dosages of FeSO₄·7H₂O and (b) the required time for the full conversion of AuNPs to AuCN with the corresponding FeSO₄·7H₂O addition.

The radical reaction nature of the FIC system was further verified via the in situ room-temperature electron paramagnetic resonance (EPR) spin-trapping technique. As shown in Fig. S2,† only one group in the EPR signal (black squares) emerged from the quaternary system containing the DMPO spin-trapping reagent, H₂O₂, Fe²⁺ and the MeCN solvent, and was made of hydroxyl radicals (DMPO–˙OH, α_N = α^β_H = 12.5 G). In contrast to the system without AuNPs, the present reaction system exhibited two groups with weak EPR signals, which can be ascribed to DMPO–˙CN (black dots) and DMPO–˙H (black triangle) species at the beginning. However, these weak signals quickly disappeared within just one minute, which is greatly different from the photo-assisted cyanation process with ˙CN radicals all along the reaction. Two causes may count for this difference: (I) compared with the photo-assisted cyanation process, the FIC process is milder, leading to a lower ˙CN concentration which is hard to detect; (II) without the assistance of UV-light, it is difficult to cleave the C–C bond in MeCN to release ˙CN under the attack of ˙OH directly. Namely, this FIC process prefers to undergo the pathway with the AuOH intermediate rather than the pathway with the ˙CN intermediate.

The reaction process was monitored by detecting the solid product at different reaction times (Fig. S3†). The reaction extent of Au to AuCN is determined via XRD where the variation of the Au/AuCN ratio suggests the conversion. Consistent with this, the colour of the suspension changes from wine red to grass green, as shown in Fig. S4.† On the other hand, the plasmon band for the AuNPs initially centred at 558 cm⁻¹ blue-shifts to 542 cm⁻¹ in the first 10 min, which correlates mostly with the size decrease of the AuNPs. As the time went on, more AuNPs were transformed to AuCN, which has a visible absorption band centered at 592 cm⁻¹, resulting in a red-shift.

Meanwhile, a notable shape transition can be seen over the period of the cyanation process. Before reaction, the AuNPs are smooth spheres supported on silica, presenting rounded edges (Fig. S5†). After 0.5 h of reaction, these rounded edges become rough with some tiny grains around the AuNPs under the attack of the ˙OH and/or ˙CN radicals. These grains are undoubtedly AuCN oligomers owing to the appearance of their corresponding XRD peaks and the remarkable colour change (Fig. S4†). As a function of the reaction time, all AuNPs are cleaved into small clusters and gradually changed to AuCN oligomers. However, it is hard to obtain the lattice fringes of AuCN oligomers owing to their easy reduction under a strong electron beam.

It is reported that ˙OH in Fenton’s reagent could quickly dissolve gold from a mechanically polished gold surface to give a much smoother surface.²⁷ Coupling this with the TEM results, we assumed that this FIC process was a surface attacking reaction, and thus the particle size may heavily affect the FIC reaction rate. To confirm this, AuNPs of 9.5 nm and 13.3 nm obtained upon calcination at 600 °C and 800 °C, respectively, were used for tests. The former fully converted to AuCN within 6 h, 2 h longer than for 5.7 nm AuNPs obtained upon calcination at 450 °C. However, the AuNPs of 13.3 nm only could be partially transformed even after 24 h of reaction with a size of unreacted AuNPs larger than 20 nm. The results of the size-dependence experiments implied that this FIC process was a size sensitive reaction: the AuNPs with a size smaller than 10 nm could be transformed to AuCN at a fast rate while the AuNPs over 20 nm transformed at a very slow rate.

The FT-IR analysis provides us with more information on the order and connectivity within MC chains. As shown in Fig. 2a, only one band (2230 cm⁻¹) corresponding to the stretching vibration of the C≡N bond (ν_C≡N) in the wavenumber range of 2400 to 2100 cm⁻¹ is found, demonstrating that the as-produced AuCN is only comprised of Au⋯CN⋯Au chains without the presence of Fe⋯CN for which ν_C≡N is centred at 2070 cm⁻¹.^9,28 The existence of a single C≡N environment in the compound is further confirmed via the observation of only one ν_Au–CN stretching frequency, which occurs at 669 cm⁻¹. Meanwhile, XPS and ICP-AES analyses were also adopted to detect the content of iron in a washed AuCN sample, and only a negligible amount of iron was observed, further ensuring the purity of the as prepared AuCN. In addition, compared with AuNPs, the Au 4f XPS binding energies of the AuCN samples shift from 87.7 to 88.9 eV for the Au 4f_5/2 signals and from 84.1 to 85.3 eV for the Au 4f_7/2 signals (Fig. 2b). The 1.2 eV up-shift of the Au 4f peaks as well as the appearance of the N 1s peak (Fig. 2b, inset) further confirm the formation of AuCN species.

Fig. 2 (a) FT-IR spectra and (b) XPS of the Au 4f and N 1s (inset) peaks of Au/SiO₂ and AuCN/SiO₂ obtained via the FIC process.

The unique structure of as-prepared AuCN was further demonstrated using extended X-ray absorption fine structure (EXAFS) and X-ray absorption near edge structure (XANES) analyses. As displayed in Fig. 3a, the XANES spectra indicate that Au is oxidized based on the intensity of the “white line” at 11 924 eV compared with Au foil, which is in consistent with the XPS results. In the Fourier transforms (r space) of the EXAFS data for AuCN (Fig. 3b), there is one prominent peak at ∼1.56 Å from the Au–C/N contribution. The other two peaks at ∼2.25 Å and ∼2.61 Å are from the Au–Au contribution, as also revealed in Au foil. The best fit for the fine structure indicates that the average coordination number is ca. 2.4 for the Au–C/N model at a distance of 1.96 Å, and ca. 3.9 for the Au–C–N or Au–N–C models at 3.14 Å (Table S1†).

Fig. 3 (a) XANES patterns and (b) Fourier transforms (r space) of the EXAFS analysis.

The similarity between Fenton-like reagents (a mixture of other transition metal ions with H₂O₂) and Fenton’s reagent further enlarges the possibilities of our FIC process. Utilizing Cu²⁺/H₂O₂ as a ˙OH provider, we successfully prepared Au_xCu_1−xCN (0.5 ≤ x ≤ 1, the x value depends on the concentration of Cu²⁺ and the reaction time). It is interesting that no CuCN phase was detected via XRD and FT-IR characterization in our sample (Fig. S6†). In addition, the fact that only one ν_C≡N frequency at 2210 cm⁻¹ is observed in the FT-IR spectrum indicates the strict alternation of the metal atoms, i.e., Au⋯Cu⋯Au⋯Cu, together with cyanide ordering.⁹ It is likely that partial Cu²⁺ was reduced to Cu⁰ NPs by H₂O₂,²⁹ which coupled with ˙CN and inserted into AuCN chains to form Au_xCu_1−xCN. A 10 h reaction between 17 mg of Cu(NO₃)₂·3H₂O, 200 μl of H₂O₂ and 60 mg of 5 wt% AuNPs/SiO₂ produced a substrate formulated as Au_0.5Cu_0.5CN, the XRD peaks of which (2 theta = 17.9, 30.2, 35.4, 36.2, 47.8, 53.9, and 57.2) matched well with the reference.⁹ AgCN and Au_xAg_1−xCN can also be obtained by wholly or partially replacing AuNPs with AgNPs or AgNO₃ (Fig. S7†).

In summary, we have demonstrated a simple way to synthesize metal cyanides and mixed metal cyanides. By using Fenton’s reagent and acetonitrile, harmful UV-light and toxic cyanide ions can be avoided efficiently, making it feasible for the large-scale industrial production of MCs. Besides, the important role that ˙OH played in the cyanation process was identified without the effects of UV-light, and provided an efficient proof for our hypothesis in photo-assisted cyanation. It is the first time that the conventional Fenton reagent was utilized in the synthesis of metal cyanides, opening a door for the application of Fenton’s reagent in synthetic chemistry. Furthermore, this facile green method provides a possibility to replace conventional metallurgical technology.

The XAFS research described in this paper was performed at the HXMA beamline of the CLS, which is supported by the CFI, NSERC, NRC, CIHR, the University of Saskatchewan, the Government of Saskatchewan, and Western Economic Diversification Canada. We are grateful for the financial support from the NSFC (21222307, 21373181, 21403197 and 21503189), the Fundamental Research Funds for the Central Universities (2014XZZX003-02), Zhejiang Provincial Natural Science Foundation of China (LY15B030009), and the China Postdoctoral Science Foundation (2014M550333 and 2015T80636).

Notes and references

1. M. J. Katz, K. Sakai and D. B. Leznoff, Chem. Soc. Rev., 2008, 37, 1884–1895.
2. O. S. Wenger, Chem. Rev., 2013, 113, 3686–3733.
3. X. Roy, L. K. Thompson, N. Coombs and M. J. MacLachlan, Angew. Chem., 2008, 47, 511–514.
4. T. D. Harris, C. Coulon, R. Clerac and J. R. Long, J. Am. Chem. Soc., 2011, 133, 123–130.
5. C. F. Shaw, Chem. Rev., 1999, 99, 2589–2600.
6. R. S. Bradley, D. C. Munro and P. N. Spencer, Trans. Faraday Soc., 1969, 65, 1920–1926.
7. H. J. Barber, A. Schonberg and A. Mustafa, J. Chem. Soc., 1943, 79–80.
8. Q. Fu, H. Saltsburg and M. Flytzani-Stephanopoulos, Science, 2003, 301, 935–938.
9. A. M. Chippindale, S. J. Hibble, E. J. Bilbe, E. Marelli, A. C. Hannon, C. Allain, R. Pansu and F. Hartl, J. Am. Chem. Soc., 2012, 134, 16387–16400.
10. S. Kroeker, R. E. Wasylishen and J. V. Hanna, J. Am. Chem. Soc., 1999, 121, 1582–1590.
11. M. Schrinner, M. Ballauff, Y. Talmon, Y. Kauffmann, J. Thun, M. Moller and J. Breu, Science, 2009, 323, 617–620.
12. T. F. Huzar, T. George and J. M. Cross, Expert Rev. Respir. Med., 2013, 7, 159–170.
13. R. H. Li, H. Kobayashi, J. W. Tong, X. Q. Yan, Y. Tang, S. H. Zou, J. B. Jin, W. Z. Yi and J. Fan, J. Am. Chem. Soc., 2012, 134, 18286–18294.
14. S. Zou, R. Li, H. Kobayashi, J. Liu and J. Fan, Chem. Commun., 2013, 49, 1906–1908.
15. J. Kim, H. J. Kim and S. Chang, Angew. Chem., 2012, 51, 11948–11959.
16. D. S. Marlin, M. M. Olmstead and P. K. Mascharak, Angew. Chem., 2001, 40, 4752–4754.
17. H. J. H. Fenton, J. Chem. Soc., Trans., 1894, 65, 899–910.
18. F. Haber and J. Weiss, Proc. R. Soc. London, Ser. A, 1934, 147, 332–351.
19. C. Walling, Acc. Chem. Res., 1975, 8, 125–131.
20. I. Hua and M. R. Hoffmann, Environ. Sci. Technol., 1997, 31, 2237–2243.
21. B. N. Aronstein, R. A. Lawal and A. Maka, Environ. Toxicol. Chem., 1994, 13, 1719–1726.
22. E. Brillas, I. Sires and M. A. Oturan, Chem. Rev., 2009, 109, 6570–6631.
23. A. D. Bokare and W. Choi, J. Hazard. Mater., 2014, 275, 121–135.
24. M. S. Lucas and J. A. Peres, Dyes Pigm., 2006, 71, 236–244.
25. Y. W. Kang and K.-Y. Hwang, Water Res., 2000, 34, 2786–2790.
26. A. Babuponnusami and K. Muthukumar, J. Environ. Chem. Eng., 2014, 2, 557–572.
27. A. M. Nowicka, U. Hasse, M. Hermes and F. Scholz, Angew. Chem., 2010, 49, 1061–1063.
28. P. J. Kulesza, M. A. Malik, A. Denca and J. Strojek, Anal. Chem., 1996, 68, 2442–2446.
29. X. Liao, R. Li, Z. Li, X. Sun, Z. Wang and J. Liu, New J. Chem., 2015, 39, 5240–5248.

---

Footnote

† Electronic supplementary information (ESI) available: Details of experimental procedures of preparation and characterization of catalysts and catalytic data. See DOI: 10.1039/c5ra01025c

---