Introduction

Lactic acid (LA) is an important organic acid which can be applied in food, pharmaceutical, and cosmetic industries [1]. Optically pure LA, especially l-LA, is the key intermediate for the production of poly-lactic acid (PLA), which is a biodegradable polymer and can be molded into various biodegradable materials [2]. Recently, with the large demand for biodegradable plastics, an alternative for petroleum-based plastics, the global interest in l-LA production is sharply increased [3, 4].

Nowadays, the commercial production of lactic acid is mainly depended on microbial fermentation due to the excellent yield, high productivity, and high optical purity of the product [5, 6]. However, at industrial scale, microbial production of lactic acid still remains two requirements: good control of pH during the fermentation process and recovery process of the acid after fermentation. Due to the low pK a value (3.8) of lactic acid, the production of lactic acid would decrease the pH of medium, and lead to the growth arrest of microbe and the decline of production rate at the end [7]. Therefore, an alkaline neutralizing reagent is often added to maintain a stable pH [8]. Calcium carbonate is the most commonly used neutralizing agent, which can remove lactic acid in time and control the fermentation pH in favor of the cell growth [9]. Many studies had reported that CaCO3 could bring a better effect than other neutralizing agents on lactic acid production [10], but this neutralization process would impose extra cost as calcium lactate requires further downstream process to be converted to lactic acid by H2SO4. In addition, a large amount of CaSO4 produced is also an environment problem in industry. Thus, developing a new way to control pH is urgent for industrial production.

So far, several methods of in situ separation and production of lactic acid have been explored, including solvent extraction, membrane-based technology, and ion exchange [11]. During extractive fermentation, the broth is mixed with an extractant and lactic acid can be extracted in time during fermentation. Keshav et al. developed a method for organic acid extraction using tertiary amines [12]. However, the tertiary amines are toxic to the cells. Moreover, an efficient extraction is more difficult than expected at relatively low concentration of lactic acid. Alternatively, membrane techniques have been integrated due to its low damage to cells. By ultra-filtration of fermentation broth using a poly-acrylonitrile membrane with a molecular weight cutoff of 20 kDa, cells were retained for next use. However, the pH control during the fermentation was still required by adding sodium hydroxide [13]. In contrast, separation and recovery of organic acids by ion-exchange adsorption had also demonstrated the advantages of high selectivity and high recovery rate, which is suitable for in situ recovery [14, 15]. An in situ separation method for fumaric acid recovery from fermentation broth was developed with a fixed bed column using IRA-900 ion-exchange resin [16]. However, its industrial application was still limited because of complex design and high process costs when using additional fixed bed column for removing the organic acid from the broth. Hence, a simper in situ recovery method was proposed to remove the produced lactic acid and control the fermentation pH during the fermentation process.

In our previous studies, Bacillus coagulans CC17 demonstrated a good performance for l-lactic acid production, which especially has potential to be used for l-lactic acid fermentation from lignocellulosic biomass [10]. However, using CaCO3 as the buffering reagent is critical for this process. There have been no previous reports about l-lactic acid fermentation by B. coagulans without adding neutralization regents. Hence, the purpose of this study is to develop a simple and economical approach for efficient in situ separation and production of l-lactic acid by B. coagulans. Several anion-exchanger resins were screened for selective adsorption and the adsorption of lactic acid under the fermentation condition was evaluated for the selected resin. Finally, an efficient in situ separation was established in l-lactic acid fermentation by B. coagulans using weak basic anion-exchange resin.

Materials and methods

Materials and resins

l-Lactic acid was obtained from Shanghai Yinggong Technology Development Co., Ltd. (Shanghai, China). All other chemicals were of reagent grade and obtained commercially.

The resins D354, D380, D941, D396, and D293 were purchased from Zhengzhou Qinshi Science and Technology Development Co., Ltd. (Zhengzhou, China). The resins D301, D315, 335, D201, and 717 were purchased from Shanghai Zhenhua Technology Development Co., Ltd. (Shanghai, China). Their structure properties are summarized in Table 1. All resins are macroporous, except for 335 and 717, which are gel-type resins.

Table 1 Structure properties of all tested resins
Full size table

All resins were first washed with distilled water and sonicated by several times to remove impurities. Then, they were washed thoroughly with 1 M NaOH to ensure complete transformation from Cl− to OH− form. Before use, all resins were washed with distilled water again to remove excess NaOH.

Microorganism and seed culture

Bacillus coagulans CC17 obtained from Nanjing Forestry University was used in all the fermentation experiments. The seed medium contained (g/L): glucose 20, yeast extract 1, corn steep liquor powder 2.5, NH4Cl 1, MgSO4·7H2O 0.2, and CaCO3 10. The starting pH was adjusted to 7.2 [17]. The seed culture was prepared as follows: a loop of cells from the fully grown slant was added into 50 mL fresh medium and incubated at 50 °C and 150 rpm for 12 h in 250 mL Erlenmeyer flasks.

Resin screening and selectivity study

For the resins screening, 1 g of wet resin was placed in a 250 mL Erlenmeyer flask and mixed with 100 mL of 10 g/L l-lactic acid solutions. Then, the mixture was incubated at 150 rpm and 50 °C for 6 h. After incubation, the solid–liquid mixture was filtered and the residual lactic acid concentration in liquid was detected as the equilibrium concentrations of lactic acid (C e). Afterwards, the adsorption resins were washed with deionized water for three times to remove the residual lactic acid on the surface, and then desorbed with 100 mL of 2 M sulphuric acid at 50 °C and 150 rpm for 6 h. The concentration of lactic acid in desorption solution was determined by HPLC. All experiments were carried out in duplicate.

The adsorption capacity, desorption capacity, and desorption ratio of the resins were calculated according to the following equations:

$${q_{\text{e}}}=\frac{{({C_0} - {C_{\text{e}}}){V_{\text{i}}}}}{W}$$
(1)
$${q_{\text{d}}}=\frac{{{C_{\text{d}}}{V_{\text{d}}}}}{W}$$
(2)
$$D=\frac{{{C_{\text{d}}}{V_{\text{d}}}}}{{({C_0} - {C_{\text{e}}}){V_{\text{i}}}}} \times 100\% ,$$
(3)

where q e is the adsorption capacity (g/g); C 0 and C e are the initial and equilibrium concentrations of lactic acid (g/L); V i is the volume of the initial sample solution (L), W is the weight of the tested wet resin (g); C d is the concentration of lactic acid in desorption solution (g/L); V d is the volume of the desorption solution (L); and D is the desorption ratio (%).

Adsorption equilibrium experiment

Adsorption equilibrium experiments were carried out in 250 mL flasks. 1 g of wet 335 resins was added to 100 mL of lactic acid solution with different initial concentrations. All samples were shaken (150 rpm) at 50 °C for 6 h to ensure exchange equilibrium. The initial and residual concentrations of lactic acid at equilibrium were determined by HPLC. All experiments were repeated in duplicate under identical conditions.

To illustrate the adsorption behaviors, the experimental data were fitted to Langmuir and Freundlich models, respectively [18,19,20]. They are expressed by the following equations:

Langmuir equation:

$${q_{\text{e}}}=\frac{{{q_{\text{m}}}{K_{\text{L}}}{C_{\text{e}}}}}{{1+{K_{\text{L}}}{C_{\text{e}}}}}$$
(4)

Freundlich equation:

$${q_{\text{e}}}={K_{\text{F}}}C_{{\text{e}}}^{{1/n}},$$
(5)

where q m is the theoretically calculated maximum adsorption capacity (g/g wet resin); K L is the Langmuir constant; and K F and n are the Freundlich constants.

Adsorption kinetic experiments

For the kinetics studies, under pH of 4, 5, and 6, 1 g of wet 335 resins were added into 100 mL of lactic acid solution with the initial concentration of 10 g/L. The pH of the solution was adjusted to the desired value with either 2 M HCl or NaOH solution. Adsorption kinetic experiments were carried out at 150 rpm and 50 °C. Liquid samples were taken at different timepoints to determine the concentrations of lactic acid.

To better illustrate the adsorption mechanism, pseudo-first-order and pseudo-second-order kinetics models were tested for simulation of the experimental data [21]. They are expressed by the following equation:

Pseudo-first-order:

$$\ln \left( {{q_{\text{e}}} - {q_{\text{t}}}} \right)={\text{ln}}{q_{\text{e}}} - {k_1}t$$
(6)

Pseudo-second-order:

$$\frac{t}{{{q_{\text{t}}}}}=\frac{1}{{{k_2}q_{{\text{e}}}^{2}}}+\frac{t}{{{q_{\text{e}}}}},$$
(7)

where q e and q t are the amounts of lactic acid adsorbed at equilibrium and at time t (h), respectively (g/g); k 1 and k 2 are rate constants of the pseudo-first-order and pseudo-second-order adsorption process, respectively (h−1).

Fermentation experiments

A traditional fermentation experiment was conducted in a 250 mL Erlenmeyer flask containing 100 mL of fresh medium. The medium contained (g/L): glucose 70, yeast extract 2.5, corn steep liquor powder 1.2, MgSO4·7H2O 0.4, (NH4)2SO4 3, KH2PO4 0.22, MnSO4·H2O 0.03, FeSO4·H2O 0.03, and CaCO3 35. Before fermentation, the culture inoculums (10% (v/v)) were added to the medium. The initial pH value of the fermentation medium was approximately 7.2 and the pH value during the whole process was controlled by CaCO3 at pH 5.0–5.5.

For in situ fermentation, the medium is the same as for traditional fermentation except to without CaCO3. 21 g of 335 wet resins were added to the medium as a pH buffering agent instead of CaCO3. Before fermentation, the initial pH was adjusted to 7.2 and the culture inoculums (10% (v/v)) were added to the medium.

All the fermentations were conducted on a rotary shaker at 50 °C and 150 rpm. The liquid samples during fermentation were collected periodically to determine the concentrations of l-lactic acid and residual sugars. The pH of solution was measured by pH meter. All experiments were performed in duplicate.

After in situ fermentation, the resin was first collected from fermentation broth by filtration, washed with water, and filtered again. Then, 100 mL of 2 M H2SO4 was mixed with the wet resin in flasks and the flasks were shaken at 50 °C and 150 rpm for 6 h for lactic acid desorption and resin regeneration. After that, the concentration of lactic acid in flasks was determined to calculate the sum of lactic acid produced in situ fermentation.

The lactic acid yield of the traditional fermentation and lactic acid yield of in situ fermentation were calculated according to the following equations.

For traditional fermentation:

$$~{D_1}=\frac{{{C_1}{V_1}}}{{{C_2}{V_1}}}.$$
(8)

For in situ fermentation:

$${D_2}=\frac{{{C_1}{V_1}+{C_3}{V_2}}}{{{C_2}{V_1}}},$$
(9)

where D 1 and D 2 are the yields of traditional fermentation and in situ fermentation, respectively (g/g); C 1 is the lactic acid concentration in the broth after fermentation (g/L); C 2 is the concentration of initial glucose (g/L); C 3 is the concentration of lactic acid eluted from the resin; V 1 is the volume of fermentation (L); and V 2 is the volume of NaOH solution.

Resin regeneration

After completion of an in situ fermentation experiment, the spent resins were regenerated by alkali–acid–alkali treatment (1 M HCl or NaOH). After that, the resin was washed with distilled water to remove excess NaOH and could be used for the next cycle. The recycling studies of in situ fermentation experiments and desorption experiments were repeated for ten times.

Analytical methods

Lactic acid and glucose concentration in all samples were determinate using a high-performance liquid chromatography system (HPLC; Agilent Technologies 1200 series, Germany) equipped with a refractive index detector. A Bio-Rad Aminex 87-H column was used at 55 °C and the mobile phase was 5 mM sulfuric acid at a flow rate of 0.6 mL/min.

Results and discussion

Resins screening and selectivity study

Since that the pKa value of lactic acid is 3.86, the production of lactic acid would acidify the medium and it is detrimental to those neutrophilic production organisms. Therefore, a neutral pH (above 5) is often needed for most fermentation process [22]. Under such neutral fermentation pH, the acid group of lactic acid would be completely dissociated and anion-exchange resins could adsorb lactic acid from fermentation broth in theory. Therefore, in this study, the static adsorption and desorption tests of various anion-exchange resins for lactic acid were conducted to identify the most suitable ion-exchange resin for the efficient separation and in situ production of l-lactic acid. The adsorption and desorption capacities, desorption ratios of the resins are shown in Fig. 1.

Fig. 1
figure 1

Adsorption and desorption capacities and desorption ratio of lactic acid for all tested resins

Full size image

Among these tested resins, D301, D354, D315, 335, D380, D941, and D396 are weakly basic anion-exchange resins, while D293, D201, and 717 are strongly basic anion-exchange resins. All skeletons of resins are acrylate or styrene, except for 335 resins, which has an epoxy structure. As shown in Fig. 1, there was no obvious difference on the adsorption ability between the weak and strong basic anion-exchange resins. However, an epoxy ion-exchange resin demonstrated a better performance than other styrene and acrylic ion-exchange resins. It implied that the skeleton structure of resin was more important than their functional groups on the adsorption ability for lactic acid. Similarly, Liu et al. reported that the hydrophilic structure of resin was in favor of lactic acid adsorption and desorption [23]. Among these tested resins, the 335 resins showed the highest adsorption capacity of 311 mg/g. Meanwhile, its desorption ratio reached the highest value of 98.2%. Moreover, it did not demonstrated any adsorption capacity for glucose even at a high concentration (70 g/L, data not shown), which suggested that using the 335 resins not only might control pH by adsorption of lactic acid during in situ fermentation, but also could obtain the high purity of lactic acid from the fermentation broth. Thus, 335 resins were selected for the further study.

Adsorption equilibrium experiment

Considering that the in situ fermentation optimum temperature by B. coagulans CC17 is 50 °C [24], the adsorption isotherms of lactic acid at 50 °C were studied to find the relationship between equilibrium adsorption capacity and equilibrium concentration. The concentrations of lactic acid in the 335 resins and in the solution are shown in Fig. 2. With the increase of feed concentration, the adsorption capacity for lactic acid increased and reached saturation status at 60 g/L. For interpretation of the adsorption experimental data, the Langmuir and Freundlich isotherms were used in this study. The Langmuir equation is frequently used to describe the adsorption behavior of a monomolecular layer, whereas the Freundlich model is an empirical equation and is widely used to describe the adsorption behavior of a monomolecular layer as well as that of a multi-molecular layer [25]. The Langmuir and Freundlich parameters are summarized in Table 2. Obviously, the Langmuir model was considered as a better model for describing adsorption equilibrium due to a higher correlation coefficient (R 2 = 0.99498). It indicated that the adsorption behavior of the 335 resins was monomolecular layer adsorption behavior. The maximum adsorption capacity of the 335 resins could reach to 402 mg/g wet resin, which was higher than most reported [26,27,28]. In most cases, the constant n in Freundlich equation refers to the adsorption intensity and the n value between 1 and 15 shows beneficial adsorption. Hence the value of n (11.3575) indicated that the adsorption on 335 resins could take place easily [29].

Fig. 2
figure 2

Langmuir and Freundlich adsorption isotherms of lactic acid on the 335 resins at 50 °C

Full size image
Table 2 Langmuir and Freundlich adsorption isotherm equations and parameters for lactic acid on the 335 resins at 50 °C
Full size table

Static adsorption kinetics

Adsorption rate is an important characteristic for adsorption process. Thus, the adsorption kinetics of lactic acid was described to understand the behavior of the 335 resins near the fermentation pH. The trends of lactic acid adsorption on the 335 resins at pH 4.0–6.0 at 50 °C are shown in Fig. 3. During the initial 1 h, the adsorption of lactic acid increased sharply with increasing reaction time. After that, the adsorption process slowed down and reached the adsorption equilibrium at about 2 h. It is well known that the extent of ionization of lactic acid molecules depends on the pH values in the solution. Thereby, the adsorption affinity is affected by the pH values. With increasing the initial pH, the adsorption capacity enhanced because of the significantly increased generation of lactate radicals. Under the experimental conditions used, the maximum of adsorption was at pH 6, which is suitable for fermentation process.

Fig. 3
figure 3

Adsorption kinetics of lactic acid on the 335 resins at different pHs at 50 °C

Full size image

Pseudo-first-order and pseudo-second-order models were employed to further illustrate the adsorption kinetic mechanism at pH 6. A summary of the results is listed in Table 3. The R2 value for the pseudo-second-order kinetic model was found to be higher than the R 2 value for the pseudo-first-order kinetic model. Moreover, the calculated qe value for the pseudo-second-order kinetic model was mainly close to the experimental data. These results suggested that kinetics in the lactic acid adsorption on the 335 resins was well described by the pseudo-second-order equation for the entire adsorption period as some reported other organic acid adsorptions on anion-exchange resins [28, 30].

Table 3 Adsorption kinetic parameters of lactic acid on the 335 resins
Full size table

In situ fermentation

Based on the above adsorption experiments, the 335 resins were chosen for lactic acid adsorption during in situ fermentation. Its high adsorption capacity made it possible to be responsible for the conversion of 70 g/L glucose in situ fermentation when using 21 g wet resin. Figure 4a shows profiles of pH value, glucose concentration, and lactic acid concentration in broth during in situ fermentation. As fermentation proceeded, the pH of medium decreased gradually due to lactic acid production. During the initial 36 h fermentation, the fermentation pH drop slowly from 7 to 5.5. After that, the fermentation pH and the lactic acid concentration in the broth tended to be stable. In the fermentation broth, lactic acid concentrations remained below 10 g/L. It suggested that the 335 resins could remove the lactate ions in time with the accumulation of lactic acid in the broth and play a good role in controlling pH at above 5.0 during the fermentation. The concentration of glucose kept decline until it was consumed completely within 72 h. It indicated that an in situ fermentation, which meant fermentation and separation simultaneously, was well done by the 335 resins and B. coagulans CC17.

Fig. 4
figure 4

Time courses of pH and concentrations of glucose and lactic acid. a In situ fermentation. b Traditional fermentation

Full size image

To prove the feasibility of in situ fermentation, the profiles of pH value, glucose concentration, and lactic acid concentration in traditional fermentation were also determined and compared with the in situ fermentation. Conventional lactic acid fermentation process was carried out using calcium carbonate as the neutralizing agent. Figure 4b shows that the pH of medium decreased rapidly during the first 12 h and maintained at about 5.0. All glucose was consumed after 72 h. More details of in situ fermentation and traditional fermentation are presented in Table 4. The amount of lactic acid production in in situ fermentation included the free and adsorbed lactic acid after fermentation. As shown in Table 4, the yield and productivity of in situ fermentation were almost the same as that of the traditional fermentation, which reached about 87% and 0.96 (g/(L h)). For B. coagulans CC17, both fermentation regimes did not affect its normal growth. Moreover, due to avoid the use of CaCO3, the subsequent downstream processing of in situ fermentation did not cause CaSO4 production, which would lead to the environmental pollution. It is concluded that the 335 resins can completely replace the CaCO3 as a pH buffer for fermentation experiment by B. coagulans CC17.

Table 4 Comparison between in situ fermentation and traditional fermentation
Full size table

Reuse of resin and repeated experiments

To assess the reproducibility of 335 resins, the adsorbed resin was regenerated and reused to the next fermentation. Ten repeated fermentation experimental results are given in Fig. 5. It is clear that lactic acid yield of in situ fermentation of 335 resins almost remained constant to above 82%. These results indicated that the 335 resins were rather stable. According to market price, the price of 335 resins is around $8 per kilogram, which is similar to CaCO3. However, the reuse of the 335 resins at least ten times obviously would save a lot of cost for production. Hence, both from economic and environmental views, 335 resins are a kind of more appropriate neutralizing reagent than CaCO3.

Fig. 5
figure 5

Reuse of the 335 resins for in situ fermentation by ten recycles

Full size image

Conclusions

In this study, an in situ fermentation method for production of l-lactic acid was established for lactic acid production with ion-exchange resin. The 335 resins were chosen for the highest adsorption capacity and selectivity from ten ion-exchange resins. After 72 h fermentation, the yield of in situ fermentation was almost the same as the traditional fermentation, which reached about 87% within 72 h fermentation. It indicated that the 335 resins could completely replace the CaCO3 as a neutralizing agent. Furthermore, ten repeat experiments showed that the 335 resins were reused at least ten times for the production of l-lactic acid. Our study provided a new way for production of l-lactic acid by fermentation. Moreover, this process was more environmentally friendly than traditional fermentation and could reduce the production cost due to much less neutralizing agent.