3. Results and Discussion
3.1 Factors Affecting P. putida F1 Biomass Preloading on CCF
Carbonized carbon fiber was chosen in this study as biofilm support considering it has large specific surface area and hydrophobic surface for hydrophobic VOC adsorption (Wang et al., 2021). In addition, hydrophobic surfaces are also considered favorable for biofilm formation as well (Farber et al., 2019). However, preloading P. putida F1 onto CCF have been proven challenging. It has been reported that different carbon sources might affect the initial biofilm formation on material surface (Dahlstrom et al., 2018). Accordingly, three types of carbon sources were tested. As can be seen from Table 1 that the biomass of P. putida F1 preloaded on CCF showed large differences when treated in MCM-based culture containing different carbon sources. Without any carbon source (thus no cell growth), the preloaded biomass (via adsorption) is 18 ± 5 mgDCW g-1(Dry cell weight per gram of CCF). However, in the presence of three different carbon sources into the MSM-based culture, the preloading ofP. putida F1 reached 35 ± 3 (Citric acid), 73 ± 6 (Sucrose), and 29 ± 4 mgDCW g-1 (Cystine), respectively. Previous research has also shown that alginate produced byP. putida F1 could promote biofilm formation (Chang et al., 2007). NaCl was selected as alginate production stimulant. However, that affected adversely the preloading of biomass, indicating the alginate might not necessarily promote affinity interactions between the CCF and cells.
Overall, sucrose promoted the highest preloading of P. putida F1 on CCF. We also observed that P. putida F1 could grow faster with sucrose than other carbon sources. So the higher preloading with sucrose could be a result of partitioning effect between the liquid media and CCF support. Apparently, higher solution phase cellular concentration will drive more cell assembling on CCF. We further tested this consideration with LB culture. As a full nutrient medium, LB media can provide P. putida F1 with sufficient nutrition and energy sources. To test the effect of cell concentration, LB-based culturing was controlled with a broader cell concentration range (with OD varied from 0.5 to 5.0). As sown in Table 2, cell concentration is positively correlated with the preloading of biomass. With OD 5.0, the preload was as high as 121 ± 8 mgDCW g-1. Biomass preloading was then conducted with OD 5.0 culturing in subsequent sample preparations in this work.
3.2 Effect of humidity on the growth of CCF-supported biofilm
Water is essential to cell growth. In the absence of a bulk liquid phase media, cell growth has to rely on moisture absorption from the gas phase. Tests were conducted to examine the growth pattern of supported biofilm under different relative humidity (RH). Figure 1A illustrates the growth curve over a five-day period with 80 % RH and 60 ppm toluene. After a one-day lag phase, the sample entered an exponential phase with a specific growth rate of 0.149 day-1 and finally had a total biomass gain of 73 mgDCWg-1. Decreased RH resulted in reduced growth and less biomass accumulation. Figure 1B~E illustrate the growth with 70, 60, 50 and 40% RH, respectively. At 70% RH, the samples experienced an extended lag period before the growth phase was reached. Under this condition, the biomass has a specific growth rate of 0.100 day-1 during the exponential phase and finally final biomass gain was 33 mgDCW g-1, which is substantially less compared to 80% RH. Decreasing the humidity further resulted in smaller specific growth rates during the exponential growth phase, which are 0.103 day-1 at 60% RH, 0.054 day-1 at 50% RH, and 0.062 day-1 at 40% RH. Besides, the corresponding total biomass gains were 33, 28, and 28 mgDCW g-1, respectively. These results indicate that water supply regulates the growth rate of P. putida F1. Similar RH dependency was also reported by Jin et al in a study of supported fungal film prepared for α-pinene treatment, which showed an optimal RH of 85 % (Jin et al., 2007).
SEM images also verified the growth of P. putida F1 on CCF. Figure 2 shows the SEM images taken for different biofilm preparations. As shown in Figure 2B, the sporadically anchored P. putida F1 cells can be observed on the surface of the CCF, while the surface of untreated samples remained smooth and clean (Fig. 2A). Biofilm growth under 80 % RH and 60 ppm toluene showed more mature biofilm development which formed a good coverage of the CCF support (Fig 2 C and D). Especially, Fig. 2D shows the texture of extracellular polymeric substances (EPS) (Nguyen et al., 2014) grown under the gaseous phase growth mode.
3.3 Effect of toluene concentration on the growth of CCF-supported biofilm
Theoretically, higher substrate concentration can promote faster biodegradation kinetics before substrate saturation is reached. However, biodegrading microbes were believed suitable for culturing at low VOC concentrations (< 1000 ppm) due to VOC toxicity (González-Martín et al., 2021). We examined biofilm growth with gaseous VOC concentrations ranged from 60 to 1200 ppm. Compared to growth with 60 ppm toluene (Fig. 1A), when the toluene concentration was increased to 300 and to 1200 ppm showed faster growth rate (Fig. 3). As shown in Figure 3, no apparent lag phase could be observed, and the specific growth rate was calculated as 0.425 day-1 at 300 ppm toluene (Fig. 3A). The growth rate at 1200 ppm decreased from that to 0.348 day-1. The final biomass gains for the two tests were 72 and 65 mgDCW g-1 for 300 and 1200 ppm toluene concentration, respectively. These results showed that increasing toluene concentration above 1000 ppm might inhibited the biomass growth, agreeing with what reported previously with aqueous culturing studies (Alagappan & Cowan, 2003; Choi et al., 2008; Reardon et al., 2000).
A comparison of growth kinetics of were P. putida F1 observed for different cultivation modes is presented in Table 3. Compared to our optimal specific growth rate of 0.425 day-1, the specific growth rate of P. putida F1 with same toluene concentration (300 ppm in liquid phase) were estimated as 1.52 and 2.12 day-1, when analyzed using Monod model and Andrews model, respectively (Abuhamed et al., 2004; Reardon et al., 2000). It is noteworthy that it is difficult to compare directly between gaseous and liquid phase growth modes though, considering differences in substrate concentration (300ppm gas phase toluene corresponds to ~80 ppm in water at sorption equilibrium). Nevertheless, we may conclude that the 0.425 day-1 specific growth rate is comparable to liquid phase growth kinetics (if estimated for 80 ppm using Monod model and data summarized in Table 3).
3.4 Long-term Toluene Biodegradation Potentials
To exam long-term biodegradation potentials, a tubular packed bed reactor (TPBR) with a continuous toluene feeding stream was built. Before packing supported biofilms in the TPBR, the reactor was tested for 24 hours to achieve a stable operation condition. The operation condition set for TPBR was 80 % RH, with a gas feeding rate fixed at 45 ml/min. Toluene biodegradation rate and CO2 production rate were monitored during the tests (Figure 4). The biomass loadings were also measured before and after the long-term operation. Compared with the initial biomass loading of 163 mgDCWg-1, the final biomass was measured as 203 mgDCW g-1 on CCF, and the final biomass gain was 40 mg (about 25% mass gain) over 10 days. A total toluene consumed of 47 mg. Therefore, the yield coefficient of biomass was calculated as 0.85 g/g, which agrees well with what have been previously reported (ranged 0.58 ~ 1.28 g/g) (Abuhamed et al., 2004; Reardon et al., 2000).
At the beginning of the operation (0 ~ 12 h), the biofilm experienced a lag phase. After that, apparent degradation of toluene took place (as reflected from the difference between the inlet and outlet toluene concentration, 12~36 h, Fig. 4). At the same time, the outlet CO2 increased from 202 ppm to 369 ppm. Degradation proceeded after 36 h but with a reduced reaction rate, as indicated by the narrowed gaps between feed and outlet concentrations (both toluene and CO2). The first few hours after the lag phase should correspond the a exponential growth phase of the biofilm, for which toluene consumption and CO2 generation rates were calculated as 2.5 mg gDCW-1 h-1 and 5.8 mg gDCW-1 h-1, respectively (based on initial biomass loading of 16%-wt). Reactions exceeding 72 h appeared to be stable throughout the operation, and should echo a stationary degradation phase. The inlet and outlet toluene concentrations were around 90 ppm and 70 ppm, respectively. In the stationary phase, supported biofilms loading increased to 20%-wt, achieved a degradation rate of 1.9 mg gDCW-1 h-1 and releasing 6.4 mg gDCW-1h-1 CO2 at the same time (indicating 100% conversion of toluene to CO2).
Overall the direct gaseous phase biodegradation showed exciting enhancement in reaction efficiency if compared to degradation rates reported previously by others. The growth phase degradation rate, measured as 2.5 mg gDCW-1h-1 in this work, is as high as 10-fold of what has been observed for liquid media growth-associated biodegradation rate of toluene (ranged from 0.17 to 0.46 mg gDCW-1 h-1 (Díaz et al., 2008; Muñoz et al., 2008). The nongrowth-associated stationary degradation phase is a comparable reaction mode to the latex coating film of P.putida F1 as reported previously by Estrada et al. (Estrada et al., 2015), which enabled degradation rates (in the rage of 0.034 to 0.321 mg gDCW-1h-1) that were also much lower than what is observed in this work (1.9 mg gDCW-1h-1).
3.4 Biofilm Growth Limiting Factors
The above observation clearly indicated that P. putida F1 biofilm could enable both growth-associated and non-growth associated degradation of toluene. The higher growth-associated degradation rate prompt us to investigate the factors that limit the duration of the exponential growth of CCF-supported P. putida F1 biofilm. As mentioned earlier, toluene toxicity may inhibit biomass growth, and there is a possibility for concentrated toluene accumulation under continuous feed operations considering that CCF could adsorb well toluene (Wang et al., 2021). In addition, insufficient nutrients (no nutrition amendment during the operation) could also be limiting.
To evaluate the effect of toluene accumulation, biofilm growth was tested in the chamber reactor under 80 % RH, but with limited toluene feed (supplied with 60 ppm toluene intermittently 6 h per day). In this case, excess toluene adsorbed onto the surface of CCFs would be released or degraded timely. As can be seen from Figure 5A, the P. putidaF1 growth pattern did not change from growth with constant toluene concentration (Fig. 1A). There was a lag phase followed by an exponential phase, and finally entered a stationary phase. That process appeared to be slower, but only slightly extended. The biomass gains ofP. putida F1 in this test was 45 mgDCWg-1, which was lower than 73 mgDCWg-1 obtained under optimal condition. This can be attributed to the less amount of toluene supply in the chamber during the growth of P. putida F1. This result indicates that toluene accumulation is not a factor that limits growth period of P. putida F1.
To test the nutrient limitation hypothesis, samples were pretreated with concentrated MSM media. Microbial growth (Fig. 5B) experienced a longer exponential phase and entered the stationary phase after 4 days. The biomass gain in this test was 136 mgDCWg-1, which was about two times greater than that we observed under standard conditions. The much increased biomass growth also become detectable with eyes (Fig. 6). As can be seen in Figure 6, there were more white accumulation appearing on doubled nutrient treated CCF than the samples normally treated. However, since the growth duration was not apparently extended, excess nutrition simply increased the overall growth rate.
In fact, previous research conducted by other researchers have also indicated biofilms did not keep growing after three or four days. One potential reason for the halt in growth could be metabolite accumulation in biofilms, as toluene degradation with the P. putida F1 could accumulate a mutagenic metabolite benzyl alcohol (Bordel et al., 2007). From our observations, this could be the primary regulating factor, but will require more systematic investigation to confirm and eventually extend the growth-associated biodegradation operations.