Alan Emsley

Kinetics and mechanisms of the low-temperature degradation of cellulose

A critical review is given of the degradation of cellulose in the low-temperature region (below about 300°C) of power transformer operation. The large number of kinetic studies, under a variety of environmental conditions from Kraft paper in insulating oil, to cotton and paper in oxygen, are considered in terms of a first-order polymer chain scission model. In many cases, the data are replotted to suit the model. A common activation energy of 111±6 kjmol−1 is calculated and it is shown that the pre-exponential factor, rather than the activation energy, is sensitive to the oxidizing nature of the environment and the susceptibility to degradation of the material. The chemical mechanisms of degradation are reviewed, and conclusions and recommendations are made regarding chemical condition monitoring and life prediction of electrical insulation.

On the kinetics of degradation of cellulose

Degradation of cellulose is important in the paper industry, where permanence determines the useful life of paper (Zou et al., 1994), in the electrical industry, where paper is used as insulation in large transformers and ultimately determines their useful life (Emsley and Stevens, 1992) and in textiles. The kinetics of paper degradation has been a topic of discussion for many years. Studies on the statistics of the process were reported by Kuhn, as early as 1930. Later, Ekenstam (1936) derived a kinetic model based on first order chemical kinetic considerations, which relates the logarithm of reciprocal degree of polymerization [ln(1=DP)] to time. Mathematically, this can be approximated to 1=DP vs time for large values of DP, the chemical equivalent of which is to assume a zero order reaction process; Hill et al. (1995) have recently shown that the same model can be derived from zero order considerations. Zou et al. (1996) have recently modelled the degradation process from first principles without any assumption of reaction order, and show that the reciprocal DP model is a special case of a more general model, which is obtained if the rate of bond scission is assumed to be constant throughout the reaction. We will show, in this letter, that degradation data can be better represented by a model which allows the rate of scission to decrease exponentially with time as the reaction proceeds. Furthermore, two fundamental constants, the initial rate of degradation and the rate of change of the degradation rate, can then be used to model change of both the DP and the tensile strength of the sample. We have shown previously that the Ekenstam model describes most of the reported data on the ageing of paper insulation in insulating oil, air, vacuum and oxygen, in the temperature range 100–200 8C (Emsley and Stevens, 1994), except at long reaction times (low DPs) where deviations occur. Zou demonstrates that it also applies to paper CELLULOSE (1997) 4, 1–5

A size exclusion chromatography study of cellulose degradation

Abstract The degradation of cellulose is an important factor in the life of electrical insulation in power transformers and in the longevity of paper archives. Traditionally, degradation has been studied by following the decrease in the viscometric degree of polymerisation (DP) in a solvent such as copper ethylenediamine (CUEN). In this paper, we report a study of molecular weight distribution (MWD) changes in cotton linters, measured by size exclusion chromatography in dimethylacetamide/lithium chloride (DMAc/LiCl) solvent, during accelerated ageing in the laboratory. The initial mono-modal distribution changes to a multi-modal distribution during ageing, but eventually returns to mono-modal as the DP of the cotton reaches a limiting value of 150–200. Such complex changes in the MWD cannot be reflected truly in any average value such as the DP, nevertheless we show that the number average molecular weight ( M n ) changes with time in very much the same way as viscometric DP and the changes can be fitted to the same kinetic model. The general trend is that the polydispersity of the material increases with ageing, as the distribution broadens, but the scatter in the data prevents any accurate trend analysis. The intermediate peak positions indicate preferential scission of molecules near the centre.

Life assessment of cable paper using slow thermal ramp methods

Abstract Rapid thermal ramp techniques are widely used to assess the properties of polymers (e.g. DTA, TGA, etc.). For certain types of processes, slow thermal ramp methods can assist the process of assessing the longer term ageing and life prediction of a polymeric material in complex environments. We will illustrate this in relation to recent work on the ageing of Kraft electrical insulation paper in insulating oil in relation to the longer term ageing of oil-filled power cables. In this case the key indicators of degradation were the degree of polymerisation (DP) and the production of gases capable of dissolving in the oil and measured by dissolved gas in oil analysis (DGA). Kinetic modelling of the change in the DP value was used to generate Arrhenius parameters and cable life prediction calculations were made for a range of potential cable operating temperatures. The DP decreases with age and increasing temperature in a complex chain scission reaction. In parallel, above 150°C there is a rapid exponential increase in the evolution of CO, CO 2 , CH 4 , C 2 H 4 and C 2 H 6 with increasing temperature. We found that the rate of degradation of the paper/oil system is dependent on the degree of containment of the ageing experiments. These results indicate that containment effects and resulting pressure effects exist in this type of accelerated non-isothermal ageing experiment and should be accounted for in practical ageing environments. The estimated life of the paper can be calculated if we assume that the Arrhenius parameters are representative of the entire degradation process and if suitable end-of-life criteria can be defined. It is also possible to calculate the average operating temperature of an electrical cable over a fixed time period, knowing the starting and finishing DP. Some of the attractions, potential weaknesses and use of the slow thermal ramp method in complex environments are also discussed.

Studies on a dicyanate containing four phenylene rings and polycyanurate blends. 2. Application of mathematical models to the catalysed polymerization process

Selected blends of bis-4-(4-cyanatophenoxy)phenyl sulphone with a commercial dicyanate, 2,2-bis(4-cyanatophenyl)propane are analysed using differential scanning calorimetry (DCS) to examine the processes of the aluminium-catalysed thermal polymerisation. Kinetic treatment of these data show that the kinetics of the formation of the bis-4-(4-cyanatophenoxy)phenyl sulphone homopolymer were fitted with just two processes, but three processes were required for the 2,2-bis(4-cyanatophenyl)propane homopolymer, for which a more complex thermogram was obtained. When considering the polymerisation kinetics of binary blends of the monomers it was necessary to select the minimum number of kinetic parameters to obtain the best fits to the data. The binary blends generally show trends in the data that reflect the monomer composition. The parameters derived from two kinetic methods are broadly in agreement; the kinetic treatment of the thermal data for 2,2-bis(4-cyanatophenyl)propane monomer suggests the presence of at least one impurity and this is supported by spectroscopic and chromatographic analyses. The latter was not observed for bis-4-(4-cyanatophenoxy)phenyl sulphone, a monomer found to be of a higher purity by chromatography. From the kinetic analysis of the thermal data (from dynamic DSC), the mathematical model predicts that, following an isothermal cure regime at 450 K, bis-4-(4-cyanatophenoxy)phenyl sulphone should reach a conversion of 90% after ca. 33 min. The empirical data for this isothermal experiment show that bis-4-(4-cyanatophenoxy)phenyl sulphone reaches a conversion of 73% after 33 min and 87% after 2 h at 450 K.

The development of controllable complex curing agents for epoxy resins Part 3.{ An investigation of the shelf life and thermal dissociation behaviour of bis(acetanilido)-tris(acetato)dicuprate(II) ;

Bis(acetanilido)-tris(acetato)dicuprate(II), [Cu2(C7H8N2O)2(CH3CO2−)3], is incorporated into MY721, a commercial epoxy resin recognised as an ‘industry standard’. Infrared spectroscopy and thermogravimetry are employed to examine the structure and thermal dissociation behaviour of the newly prepared complex. Multivariate analysis techniques (e.g. principal components analysis) are used to examine the spectral data to corroborate the dissociation temperature. The shelf life and cure characteristics of the commercial epoxy/curing agent formulation are determined using differential scanning calorimetry and vibrational spectroscopy.

A solid-state NMR study of cellulose degradation

A series of laboratory-aged transformer insulating papers were investigated using solid-state NMR spectroscopy. Carbon-13 CPMAS, and proton MAS experiments were carried out along with static proton relaxation (T1, and T1ρ) and free induction decay (FID) measurements. Some proton CRAMPS and proton-carbon-13 correlation (WISE) experiments were also undertaken. A change in the proton T1 and FID with ageing was detected. No detectable change was found in the proton T1ρ. Some amorphous cellulose was detected in the carbon-13 spectrum. There was, however, no evidence for a substantial change in the nature of the cellulose with ageing. The carbon-13 spectra from some aged samples showed signals not present in the spectrum from an unaged sample. This was taken to be evidence of chemical degradation. Proton MAS and the WISE exeriment gave some information about the nature of the water in the sample.

Ultra-Accelerated Thermal Ageing Tests

Isothermal testing of polymer ageing requires long-term exposure if realistic results are to be obtained, which means a time-consuming and hence expensive exercise, which may have to be repeated many times over many different conditions. Rates of ageing can be accelerated by increasing the temperature of tests or by using a more aggressive environment, but then results must be extrapolated to working conditions, assuming that the same mechanisms still apply. In this paper, we demonstrate that a slow thermal ramp technique, similar to thermo-gravimetric analysis, allows the qualitative and, in some cases, quantitative assessment of ageing to be made in a relatively short time, 15 to 20 days, compared to up to 2 years for isothermal tests. We will illustrate the method in relation to recent work on the ageing of Kraft electrical insulation paper in insulating oil and compare results from isothermal and thermal ramp experiments. In this case, Degree of Polymerisation (DP) was used as a key indicator of degradation.