**3. Mechanism of vanillin formation from lignin and the roles of organic cations**

Bu4N+ and the complex cations between Na<sup>+</sup> and the polycyclic ethers are found to facilitate the vanillin production by alkaline aerobic oxidation of lignin. However, the mechanisms underlying the improved vanillin yield caused by the presence of the organic cations are not clear. We, thus, investigated the role of organic cations (Bu4N+ and the crown ether-based complex cations) in the vanillin formation mechanisms, by analyzing the behaviors of lignin model compounds **LM1** and **LM2** shown in **Figure 3** under the alkaline aerobic oxidation conditions. These models carry a β-O-4 linkage, which is the most abundant linkage in lignin, but are different in that **LM1** has the phenolic OH group, while **LM2** does not.

**LM1** with the phenolic OH group was examined as a model for the phenolic end group of lignin [16]. As shown in **Figure 3**, an enol ether intermediate **EE** is formed in the initial stage of the oxidation via the formation of a quinone methide **QM** structure from **LM1** and the following elimination of Cγ position as HCHO (Pathway A1 and A2). The second step involves the oxidation of the Cα = Cβ moiety of **EE** finally leading to vanillin (Pathway A3). The mechanisms in the oxidation of **EE** are not clear at the moment, but vanillin formation from **EE** via a dioxetane intermediate was suggested in several studies on O2 pulp bleaching conditions [40, 41]. For the effect of Bu4N+ in the TBAH medium, Bu4N+ stabilizes **EE** probably by coordinating with the side chain of **EE**, which results in slower vanillin formation from **LM1** in TBAH

*Depolymerization of Native Lignin into Vanillin, Vanillic Acid, and Other Related Compounds… DOI: http://dx.doi.org/10.5772/intechopen.112090*

#### **Figure 3.**

*Reaction pathways involved in alkaline aerobic oxidation of* **LM1** *(A) and* **LM2** *(B) [16, 33, 39].*

compared to that in NaOH. It is, therefore, likely that Bu4N+ does not have positive effects on the vanillin production from the phenolic end of lignin.

Experiments using **LM2**, a model for the middle β-O-4 phenylpropane unit of lignin, have suggested the following reaction mechanisms [16, 39]. First, the cleavage of the β-ether of **LM2** via the neighboring group participation of Cα-OH occurs to give a veratryl glycerol intermediate **VGL** (Pathway B1). This intermediate is then oxidized to an α-carbonyl compound, which is in equilibrium with the corresponding γ-carbonyl compound through the keto-enol tautomerization (Pathway B2). A retro-aldol reaction proceeds from the γ-carbonyl compound, leading to the formation of veratraldehyde (**VAld**) via Pathway B3. On the other hand, another retro-aldol pathway starting from the α-carbonyl is less significant [39]. There are two possible fates of **VAld** with one being disproportionation to produce veratric acid (**VAcid**) and veratryl alcohol (**VAlc**) (Pathway B4) and the other being conversion of the 4-OMe group to OH to produce vanillin (Pathway B5).

Comparison of the degradation behaviors of **LM2** in the TBAH and NaOH system revealed that TBAH promoted the cleavage of the β-ether in the initial stage of the reaction (Pathway B1) and increased the selectivity of the vanillin production from **VAld** in the final stage (Pathway B5). The former promotion of the β-ether cleavage can be explained by the fact that the OH− activity in the aqueous TBAH solution is higher than that of the aqueous NaOH solution, even when their OH<sup>−</sup> concentration is the same. On the other hand, the latter promotion of vanillin formation from **VAld** is due to the interaction of Bu4N+ with **VAld** in the aqueous TBAH solution, as indicated by the <sup>1</sup> H NMR analysis of **VAld** in the presence of Bu4N+ salt [16]. In

other words, **VAld** is in the cage made of Bu4N+ in the aqueous TBAH solution, which inhibits the progress of the disproportionation (Pathway B4). This results in relative increase in the selectivity of the vanillin forming pathway B5.

The influence of the complex cations on the reactivity of **LM2** is confirmed by evaluating the additive effect of crown ethers (**18C6**, **15C5**, and **12C4**) on the reactivity of **LM2** in the aqueous NaOH solution. When the complex cations are present in the reaction system, a decrease in the yields of **VAcid** and **VAlc** is observed, along with an increase in vanillin yield, similar to the previously confirmed Bu4N+ system. Thus, it is inferred that both complex cations and Bu4N+ essentially increase the vanillin yield through the same mechanism, namely, the suppression of the disproportionation pathway of **VAld** due to the cage effect. Interestingly, in the case of the non-cyclic ethers such as **TEG** and **TRG**, and 1,4-dioxane, a cyclic ether with a much smaller ring size, no significant increase in vanillin yield is observed in the model experiments using **LM2**. This suggests that the formation of complex cations shown in **Figure 1**, and the suppression of the disproportionation pathway (Pathway B4) are closely related. In addition to the similar effects of Bu4N+ and the complex cations as mentioned above, the complex cations exhibit several specific effects that are not observed with Bu4N+ [32]. For instance, in the case of the complex cation systems, the inhibition of the oxidative degradation of the **VGL** side chain to the aldehyde group is observed probably by the stabilization of **VGL** by its interaction with the complex cation.

The above phenomena observed for **LM1** and **LM2** suggest that vanillin formation from middle phenylpropane units in lignin is promoted in the presence of the organic cations, with the reaction at the phenolic end being rather suppressed. In other words, the formation of vanillin from the middle unit plays an important role in the formation of vanillin under the alkaline aerobic oxidation conditions. **Figure 4** shows a possible pathway for vanillin formation from the middle unit of lignin based on the study of the reaction mechanisms of **LM2**. First, the β-ether of the middle β-O-4 unit undergoes alkaline hydrolysis to form the glycerol end (pathway І). The glycerol end is oxidized to a Cα-aldehyde end (pathway Π), and finally, vanillin is released from this end (pathway Ш). In an actual lignin molecule, the middle units are considered to be more abundant than the phenolic ends. It is, thus, likely that the suppressing effect of Bu4N<sup>+</sup> on the vanillin production from the phenolic end is canceled out, and the overall vanillin yield was increased in the organic cationcontaining media.

#### **Figure 4.** *Vanillin formation from a β-O-4 middle unit of lignin polymer [16].*

*Depolymerization of Native Lignin into Vanillin, Vanillic Acid, and Other Related Compounds… DOI: http://dx.doi.org/10.5772/intechopen.112090*

The model compounds **LM1** and **LM2** discussed so far are designed to mimic the β-O-4 structure in lignin. However, the compound has a much more simplified structure than the actual lignin. Therefore, when applying the reaction pathway based on the degradation behavior of **LM2** (**Figure 3**) to actual lignin molecules, careful consideration is required. When comparing the vanillin production behavior from **LM2** and wood flour, a certain correspondence can be recognized between the oxidation behaviors of actual lignin and the model compounds in that the increase in vanillin yield from the wood flour observed in the presence of organic cations is reproduced even in the case of the model compound. Thus, it is likely that the suppression of the disproportionation by the cage effect of the organic cations plays a certain role also in the vanillin production from actual lignin. In the case of actual lignin, however, **VAld**, which was produced from **LM2**, never forms, and instead the Cα-aldehyde end group shown in **Figure 4** is formed as a vanillin precursor. The rates of vanillin elimination from this Cα-aldehyde end and the conversion reaction of **VAld** to vanillin (conversion of the 4-position methoxy group to a hydroxy group) are expected to differ significantly, which may affect the degree of suppression of disproportionation by the cage effect. Also, in the Cα-aldehyde end linked to the actual lignin side chain, more complex side reactions not observed with **VAld** might proceed. According to these ideas, future studies using model compounds that more accurately mimic the actual Cα-aldehyde end are needed to more precisely elucidate the mechanism of increased vanillin yield by the cage effect.
