Optimization Of An Ion Chromatography-Based PHA Measurement Technique To Support BPR Operation

INTRODUCTION
Polyhydroxyalkanoate (PHA) is a chemical produced and used by microorganisms during biological phosphorus removal (BPR) to power uptake of phosphorus (Satoh et al., 2016; Liu et al., 2018). PHA is produced from volatile fatty acids (VFA) during the anaerobic stage of BPR. Polyhydroxybutryate (PHB) and polyhydroxyvalerate (PHV) are the two major species of PHA observed in activated sludge systems. It is useful to be able to detect this chemical because PHA has been linked to BPR stability in basin and laboratory scale studies (Schauer et al., 2018; Liu et al., 2018). Clean Water Services is an advanced wastewater treatment facility that uses BPR to achieve permit compliance with minimal use of chemical effluent polishing. We have determined biomass PHA content to be an operational parameter relevant to BPR stability, and so have chosen to develop an in-house method for PHA detection. An in-house method has the benefit of providing data quickly for use in BPR process management. The most widely favored method for PHA detection is gas chromatography (GC; Tan et al., 2014; Koller et al, 2015). While highly valued for its detection sensitivity, GC analysis of PHA is time and labor intensive, and requires the use of hazardous and volatile solvents during the digestion process. In brief, this method requires that an activated sludge sample be preserved with formaldehyde, freeze dried/lyophilized overnight, then digested at high temperature using chloroform and methanol for several hours while shaking (Tan et al, 2014). The risk of explosion during this stage is significant due to the high temperatures, volatility of the solvents, and agitation of the liquid (Koller et al, 2015). For these reasons, pursuing an alternative method for PHA detection was deemed advisable. Hesselmann et al. published an alternative to the GC method using ion chromatography (IC) for PHA preservation, digestion, and detection in 1999. While shown to be accurate and capable of quantitatively measuring and differentiating between the monomers that comprise PHB and PHV (important, since speciation of PHA is highly influenced by VFA feedstock), there have been few further developments reported in the literature since. We chose to optimize this method to develop and validate an in-house IC based protocol for PHA detection. This method is safer, faster, and requires less specialized sample handling than the currently used GC method.
METHODS
We optimized the protocol set forward by Hesselmann, et al. (1999), making alterations to further improve PHA preservation and recovery. A condensed final protocol for sample preservation is presented in Table 1. Table 2 describes the instrumentation and equipment parameters for the IC measurement of PHA, while Table 3 lists QA/QC parameters.
RESULTS
Method optimization We made alterations to the method reported by Hesselmann et al (1999) in order to improve reproducibility and reduce error between replicates. The preservation and digestion steps were optimized by measuring PHA recovery from biomass using both biological and technical replicates. The most substantial alteration to the Hesselmann method was addition of a drying step during sample preservation, as opposed to simply pipetting as dictated by the original method. We determined that decanting and drying samples to remove the supernatant resulted in better PHA recovery and less error than pipetting. The improvement to PHV recovery was particularly dramatic, increasing by 30% (Figure 1). This also improved repeatability, with error between replicates declining from approximately 10% (PHB) and 20% (PHV) to less than 1% for both PHA species (Figure 2).
PHV reference standards were not commercially available when the original method was published, requiring calculation of PHV from a PHB reference standard. However, a PHV standard is now available, allowing us to directly calibrate to PHV monomers. These alterations made to the IC method enabled us to improve the calibration range from Hesselmann's reported 1-16 mg/L to 0.05-5.0 mg/L. This gave us not only a lower detection limit but also better sensitivity, resulting in greater precision between replicates. Method validation via VFA/PHA loading The IC method to measure PHA was successful, producing high accuracy data with low variability between replicates. This was demonstrated by loading activated sludge biomass with PHA, then digesting and measuring the amount of PHA recovered. While not all VFA is expected to be recovered as PHA since we were working with a mixed microbial biomass (including glycogen accumulating organisms, as well as phosphorus accumulating organisms that do not store VFA as PHA), high PHA recovery would indicate good digestion efficiency of the method. To show this, we added precise amounts of VFA as COD to activated sludge bioreactors and allowed VFA uptake to occur under anaerobic conditions. Once uptake was complete, we measured the amount of PHA produced within the cells and compared it to original VFA added to estimate digestion efficiency and PHA recovery from biomass (Figure 3). PHA recovery was excellent, averaging 82%. Error between PHA replicates was low, highlighting the reproducibility of the IC method (Figures 4 and 2). When this experiment was repeated with varying acetate concentrations (40 mg/L and 80 mg/L), PHA recovery was doubled in the reactor with twice the amount of acetate, confirming extraction efficiency at a scale expected based on VFA input (Figure 5).
CONCLUSIONS
We have optimized and validated a safe, reliable method for PHA detection using an IC. The IC method returns PHA values that are consistent with expectations based on VFA loading. As the method is optimized to achieve maximum recovery of PHA from activated sludge mixed liquor, these data are operationally relevant to BPR stability considerations. Our results show that this method is capable of distinguishing quantitatively between PHB and PHV, with low relative error between replicates. We demonstrated high recovery of PHA from activated sludge biomass, indicating good digestion efficiency and instrumentation sensitivity. This method improves upon the originally published protocol, as demonstrated by our ability to improve the calibration range by two orders of magnitude from Hesselmann's reported 1-16 mg/L to a much more sensitive 0.05-5.0 mg/L. This allows our dilutions to generate more accurate data for high PHA samples than the previous range. The implementation of this updated protocol will make measurement of PHA and its relation to BPR stability and process control more accessible to facilities with analytical capacity. The full paper will include expanded data pertaining to method optimization and replicated experiments, as well as demonstration of successful method application to determining PHA dynamics at different stages of the BPR process, and at different stages of BPR stability.