Water and alkali extracted steam exploded yellow poplar fiber (log R₀ 4.25) was purified through a multi-stage process. The initial stage consisted of peralkanoic acid treatment, either formic or acetic acid, at atmospheric pressure. Subsequent bleaching was achieved through treatment with alkaline peroxide in several stages. Performic acid experiments were performed by researchers at the Finnish Pulp and Paper Research Institute: K. Poppius, I. Tuominen and J. Sundquist. Peracetic experiments were conducted at Virginia Tech.

The alkanoic acid stage process parameters examined included time (3 and 6 hours primarily), temperature (40 and 60° for the peracetic trials, 80° for the performic trials), acid concentration (25, 50, and 80% for the peracetic trials), and initial peroxide charge (10 and 20 % for the peracetic trials; 2,5, and 10% for the performic trials).

The bleached fiber was analyzed on the basis of brightness (diffuse reflectance of hand sheets at 457nm), Kappa number, molecular weight (carbanilation followed by GPC), and carbohydrate composition (complete hydrolysis followed by HPLC). Peroxide concentration was monitored through iodometric titration.

The results indicated that the purified fiber may be useful as a microcrystalline cellulose. The purification process occurred in three phases: lignin activation, followed by dissolution, followed by alkaline peroxide bleaching. Brightness correlated well with Kappa number. Molecular weight loss (as weight average) averaged 43%. Glucose purity of 93% was realized for the peracetic trials with residual lignin of approximately 1%. Residual hemicellulose content was approximately 1.5%. Final brightness levels of nearly 90% MgO were attained by the performic acid samples.

In terms of percent gain in brightness per percent peroxide consumed, higher temperature resulted in greater efficiency. An alkaline extraction (no peroxide) following the initial acid stage improved efficiency. Initial peracetic peroxide charge of 20% was inefficient compared to an initial charge of 10%.

Molecular weight loss was affected by water concentration of the acid stage, but unaffected by an increase in the acid stage temperature from 40° to 60°. Molecular weight loss appeared to increase with acid stage peroxide charge up to 10%, but 20% charge did not result in greater losses.

Further research using higher temperatures and a sulfuric acid catalyst are suggested.