There is a rather good correlation between different analytical methods used for the analysis of native lignin preparations, both in intralab (Evtuguin et al., 2001) and in interlab studies (Sakakibara, 1991; Capanema et al., 2004; Zhang and Gellerstedt, 2007; Balakshin et al.,

2008) . An exception is the 31P NMR analytical methodol­ogy of native lignins which yielded ca. 30% lower numbers for aliphatic hydroxyls (Pu et al., 2011) compared to other methods (Sakakibara, 1991; Balakshin et al., 2008).

A rather good correlation between various methods for technical lignins analysis has been also reported (Faix et al., 1994; Cateto et al., 2008). However, a compre­hensive review of published analytical data leads to a much less optimistic conclusion. Table 18.3 shows that significant variability in the structure of the same tech­nical lignins can be observed when these lignins are analyzed by independent methods, in contrast to what is seen with the analysis of native lignin preparations. This deviation might be caused by the interference of specific lignin moieties generated during the technical process on the results of each specific analytical method which, as a rule, was developed and validated for the analysis of native lignin preparations.

Various wet chemistry techniques for the analysis of lignin functional groups have been comprehensively reviewed previously (Lin and Dence, 1992; Zakis, 1994). Therefore, we will focus our discussion on major NMR spectroscopic techniques for the analysis of technical lig­nins, 31P and 13C NMR, as well as advanced NMR methods, which have received less attention.

As 31P NMR spectroscopy of derivatized lignins be­comes one of the most common techniques for lignin anal­ysis, it is important to critically evaluate the method in a comprehensive manner. There are two main

modifications of the 31P NMR lignin analysis. Originally,

2- chloro-1,3,2-dioxaphospholane (31P-I method) was sug­gested (Archipov et al., 1991) as the derivatizing agent in this method. Later, 2-chloro-4,4,5,5-tetramethyl-

1,3,2- dioxaphospholane (31P-II) was reported

(Kostukevich et al., 1993; Granata and Argyropoulos, 1995) to provide better signal separation and it is currently used as the major 31P method for lignin analysis. Although good agreement has been reported between the results obtained with these two derivatization re­agents (Granata and Argyropoulos, 1995), data reported later did not confirm this observation (Tables 18.3 and

18.4) . The results obtained with the 31P-II method tend to underestimate results compared to the data generated by all other analytical methods. The 31P-I method tends to report significantly higher amount of aliphatic and total hydroxyl groups if compared to the data obtained by the 31P-II methodology (Table 18.3), in contrast to the conclu­sions drawn in the original validation work (Granata and Argyropoulos, 1995).

The significantly lower numbers reported by the 31P-II NMR analysis, as it is compared to other analytical methods, in the structural analysis of lignins might be explained by the incomplete lignin derivatization with the phosphorylating reagent PR-II possibly due to steric hindrance of the bulky reagent containing four t-butyl groups. The use of PR-I yields apparently more quanti­tative results. However, the signal resolution in 31P-I is not high enough as it can clearly be seen in the publica­tion by Akim et al. (2001). In this publication, the signals of primary hydroxyls and 5-substituted phenolic hy­droxyls are heavily overlapped. The main conclusion derived from this observation is that even when the res­olution of a resonance signal is formally acceptable (Argyropoulos, 1994), one cannot conclude that the resolved signals are reporting the correct values.