Ethanol Production Monitoring Using Ion Exclusion HPLC

By Michael McGinley and Jim Mott | June 02, 2008
The current process for generating ethanol from biological sources relies on the use of amylase enzymes to break down complex starches into simple sugars, followed by yeast fermentation to convert the sugar to ethanol. Regular monitoring of the process by high-performance liquid chromatography (HPLC) allows operators to track the breakdown of starches to simple sugars as well as monitor ethanol production after yeast has been introduced. Organic acids are also monitored in the HPLC run, allowing operators to assess if microbial contamination is affecting the fermentation process and if remediation steps such as antibiotic addition are necessary to maximize ethanol yield.

The HPLC method used for ethanol monitoring, ion exclusion chromatography, uses several different separation modes (gel filtration, ion exchange and reversed phase) to separate compounds of interest. This separation method has been around for more than 20 years, yet still remains the most popular for fermentation monitoring due to the ability to separate different classes of compounds (sugars, organic acids and alcohols) all in one chromatographic separation.

The method is a simple and rugged isocratic method using a dilute acid mobile phase. However, some sample preparation of the fermentation broth is required to ensure good chromatographic performance and a reasonable column lifetime.

Many ethanol producers are expanding their operations by adding additional fermentors. In order to continue using existing HPLC equipment for monitoring of more fermentors, increased analytical throughput is needed. The current ion exclusion HPLC method is a multimodal separation so increasing throughput can be a challenge. However, some minor changes can be implemented that can reduce the analysis time by up to 50 percent, from a 24-minute to 12-minute run time.

This article discusses the basic principles of the ion exclusion HPLC analysis, along with a review of sample preparation methods to ensure proper cleanup and minimize instrument down time will be reviewed. Also discussed are ways to reduce analysis time and increase throughput.

Materials and Methods
Analyses were performed using a Shimadzu LC-20AT LC system equipped with a SIL-10AF autosampler, degasser and a RID-10A RI detector. Data were collected using Class-VP Version 7 software.

Figure 1

An ethanol fermentation standard was run on a 300 mm by 7.8 mm column. Note the excellent separation of all the components of interest. 1) Dp4+, 2) Dp3, 3) Maltose, 4) glucose, 5) lactic acid, 6) glycerol, 7) acetic acid, 8) ethanol

Figure 2

The bioethanol fermentation standard was run on a 150 by 7.8 mm Rezek ROA column. Note the limited resolution of the early eluting oligosaccharide peaks in the standard. In early fermentation monitoring such peaks may not be resolved. 1) Dp4+, 2) Dp3, 3) Maltose, 4) glucose, 5) lactic acid, 6) glycerol, 7) acetic acid, 8) ethanol.

Various dimensions ofPhenomenex Rezex ROA columns were used (150 millimeter (mm) by 7.8 mm and 300 mm x 7.8 mm). Guard columns were Phenomenex SecurityGuard Carbo-H+ 4 mm by 3 mm cartridges. Aqueous mobile phase (0.005 N sulfuric acid in water) was purchased from Chata Biosystems Inc. By definition, a 1 N solution contains one equivalent per liter. Several samples from various fermentation time points were generously provided by ICM Inc. The ethanol HPLC testing standard was obtained from Midland Scientific.

Crude samples from fermentation time points were filtered using a 0.20-micron Phenex-RC syringe tip filter. Filtered aliquots of 10 microliters were injected on HPLC operating at a flow rate of 0.6 milliliters per minute. The HPLC column was heated to 65 degrees Celsius (149 degrees Fahrenheit). Run time was 24 minutes for the 300-mm column and 14 minutes for the 150-mm column. SecurityGuard cartridges were regularly changed every 100 runs or whenever static increased 10 percent above initial values. Fifty percent methanol was used in the autosampler needle wash to avoid bacterial contamination.

Results and Discussion
An HPLC run of a fermentation standard using the Rezex ROA 300 mm by 7.8 mm column is shown in Figure 1. Note that the early eluting peaks (Dp4+, Dp3, maltose and glucose) represent the different degrees of polymerization of the various saccharides present in the sample. Monitoring of these peaks during early time points of the fermentation run gives operators a good indication as to the progression of the various amylases used to break down starches to simple sugars, and dictate when yeast is added to the fermentor to start generating ethanol. The later eluting peaks (lactic acid, glycerol, acetic acid and ethanol) represent the organic acids and alcohols generated during the fermentation. Monitoring of these peaks gives an operator an indication as to the fermentation endpoint and reveals when bacterial contamination is severe enough to warrant addition of an antibiotic to limit bacterial byproducts that may inhibit ethanol production.

The standard in Figure 1 shows idealized separation of compounds. Early in the fermentation, the oligosaccharide and saccharide peaks can be so abundant that resolution between components is reduced. Later in the run, as sugars are converted into ethanol, resolution of the saccharides is more in line with what is seen in the HPLC standard. Conversely, early in the process, ethanol, glycerol and organic acid peaks often are below detection limits and increase as fermentation progresses.

Resolution of key components will also tend to decrease over time as sample contaminants build up on the column. The key to maintaining resolution and increasing column lifetime is using a guard column system such as the SecurityGuard cartridge system.

One method for reducing HPLC run time is by reducing the length of the column used. An example is shown in Figure 2 where the fermentation standard is run on a 150 mm by 7.8 mm (half the length of the typical 300 mm column used). As expected, the run time using a shorter column is significantly reduced from 24 to 13 minutes. If one looks closely, while the resolution of the late eluting organic acid and alcohol peaks are acceptable, the early eluting oligosaccharide and saccharide peaks are significantly reduced. Since the separation of oligosaccharides is based primarily on a gel filtration mechanism, there is a limitation on how much the column length can be shortened and still maintain resolution of key saccharide peaks.

While the analysis of early time points from a fermentation run may need the increased resolving power of the longer 300 mm by 7.8 mm column, it may be practical to use a 150 mm by 7.8 mm column for later time points, where saccharide peaks are smaller. This is most practical in a larger operation where multiple fermentors are used and multiple HPLCs might be used for monitoring. Alternatively, for sites where only one HPLC is used, a column-switching valve might be employed to switch to the shorter column later in a fermentation run.

Simple steps such as using guard columns and filtering samples can greatly increase the reliability of the method as well as improve column lifetime (thus reducing analysis cost). For larger operations, shorter columns can be used in some circumstances to reduce analysis time for monitoring, potentially reducing the need for additional analytical equipment.

Michael McGinley is a biochromatography product manager at Torrance, Calif.-based Phenomenex Inc. Jim Mott is the senior technical support specialist in Shimadzu Scientific Instruments Inc.'s Midwest regional office in Lenexa, Kan. Reach him at or (913) 888-9449.