Physical characterization and kinetic studies of Zn (II) biosorption by Morganella morganii ACZ05

Through this investigation, we establish the mechanism and physical characterization of zinc (II) sequestration by Morganella morganii ACZ05 strain, which was isolated and characterized from soil polluted by ef ﬂ uents from electroplating industries. As far as we know, there is very little literature concerning zinc biosorption using an environmental strain of Morganella morganii . The SEM analysis shows the dark porous gaps in the aggregated cell-matrix of test bacterial biomass which is inferred as water channels usually seen in bio ﬁ lms, as compared to metal-unexposed control. Morganella morganii is not known to produce bio ﬁ lms unless in the rare nosocomial conditions. Here, SEM analysis shows the production of bio ﬁ lms after exposure to zinc (II) at 500 ppm, which has not been previously reported. EDX analysis of bacterial biomass also speci ﬁ ed the sorption of zinc (II) by the bacterial cells and the presence of new peaks for zinc in contrast to control. Both XRD and FTIR analysis observations strongly implicate the potential of physical adsorption as a mechanism for heavy metal resistance. Analysis of the cell surface by Atomic force microscopy and examination of the topography revealed cell aggregation occurs during bio ﬁ lm production after zinc biosorption. Unlike other reports, regular models such as Langmuir isotherm and Freundlich isotherm were found insuf-ﬁ cient to explain the physisorption of zinc (II) metal ions on the complex multicomponent adsorbents such as the exopolymeric surface of the bacterial cells. However, adsorption kinetics of zinc (II) to the bacterial biomass was most effectively elucidated by a pseudo-second-order kinetic model, suggesting a certain kind of chemisorption that requires further study. SEM-EDX and AFM The isotherms studies – ﬁ t of pseudo-second order for biosorption experimental data suggests the involvement of cellular mechanisms for the activity of binding sites of metal ions and other chemical sorption processes. Biosorption, chemisorption,

the numerous adsorption isotherm models developed about metal sorption. But they can provide an insight into the involvement of various modes of sorption such as physisorption and chemisorption of metal ions on the surface of the biosorbent (Kotrba 2011).
Similar studies indicate that heavy metal biosorption can be modeled using Langmuir and Freundlich isotherm model. These models are uncomplicated, dependent on two parameters concentration and time, which points us towards the equilibrium reached under experimental conditions. Langmuir model stipulates that the surface of the bacterial biomass has different binding sites that possess affinity towards certain metal ions but the adsorbate molecules do not interact with each other probably creating a monolayer. On the contrary, the Freundlich isotherm model supposes that occupancy of adjacent sites for the binding of metals and the bond energy between the adsorbent and adsorbate are related (Lu et al. 2006;Gabr et al. 2008;Joo et al. 2010;Huang & Liu 2013).
The sorbate and the biosorbent exist are pushed forward in an adsorption reaction to reach equilibrium, in an aqueous phase, and determines the sorbent's maximum uptake ability. The kinetics data of the sorption provides us optimum of all parameters such as sorbate removal efficiencies, adsorbent (biomass) concentrations, pH, and temperature. These contribute significantly during the development of a wastewater treatment plant or a setup (Derco & Vrana 2018). During sorption studied in batch reactors conditions, kinetic data are characterized by verifiable models i.e., pseudo-first and -second-order models (Kotrba 2011).
The metal ion and biosorbent surface interact in two phases due to their biphasic nature. 90% of the biosorption occurs in the initial phase which is rapid and plateaus for a fixed time. The second phase of biosorption occurs rather slower lasting up to 4 hours. For example, in a study by Çeribasi & Yetis (2001), when Phanerochaete chrysosporium is exposed to varying concentrations of nickel and lead, maximum adsorption was attained usually in the first 30 minutes, whereas it took 3 hours to reach equilibrium. Lu et al. (2006) reported that biosorption kinetics and equilibria of lead (Pb), cadmium (Cd) and copper (Cu) ions by Enterobacter sp. The results showed that the biomass of Enterobacter sp. was an efficient adsorbent of lead (Pb), cadmium (Cd) and copper (Cu) ions suggesting good recyclable of the biosorbent. For analysis, kinetics of heavymetal adsorption by Enterobacter sp. and the model seemed to have an outstanding prediction of the experimental information. Likewise, Morganella showed better metal removal efficiency. Morganella is a type of gram-negative microbes belonging to the family Enterobacteriaceae, members of which were formerly planned as a candidate in bioremediation of chromium (VI), Lead (II) and uranium (VI) (Martins et al. 2010;Kumar et al. 2020;Princy et al. 2020).
Our study determined the biosorption of Zn (II) by Morganella morganii ACZ05. Equilibrium dynamics and kinetics of the biosorption were investigated under batch reactor conditions by atomic absorption spectroscopy. Isotherm modeling and kinetics models were fitted with experimental data to determine the best fit. XRD, FT-IR, SEM-EDX, and AFM were used to examine the changes in the biosorbent surface post-sorption, and mode of biosorption.

Media used and growth conditions
A single Zn (II) resistant bacterial strain, Morganella morganii ACZ05 was characterized from soil near electroplating industries situated in the Jaihindpuram area of Madurai. 100 μl of the soil and effluent samples were plated onto sterile nutrient agar plates after diluting them to 10 À6 . The nutrient agar media containing; (Peptone 5 g/l), sodium chloride (5 g/l), HM peptone B (1.5 g/ml), yeast extract (1.5 g/l), agar (15 g/l) at 28°C. The pH of the growth medium was adjusted to 7.2 by the addition of 1 M KOH prior to autoclaving. The bacterial isolates were then grown for 24 hrs at 37°C. A total of ten zinc metal-resistant bacterial strains were isolated. Among the ten strains, two strains were able to tolerate 1,000 ppm concentration of zinc (II) metal. From which the only one highest tolerant bacterial strain was chosen for the study.
The bacterial isolates so obtained, were tested for tolerance in different concentrations (250 mg L À1 , 500 mg L À1 , 750 mg L À1 and 1,000 mg L À1 ) of zinc (II). After incubation at 37°C for 24 h, the bacterial strain which could tolerate 1,000 mg L À1 zinc was selected and identified as M. morganii ACZ05 (GenBank Accession No: KJ830754).

Preparation of metal stock solutions
Zinc sulfate (ZnSO 4. 7H 2 O) was dissolved in double distilled water to get a standard solution of zinc (II) (1,000 mg L À1 ). Atomic absorption spectrophotometer (ELICO Technologies Pvt. Ltd, India) was used to measure the zinc (II) metal ion concentration in all samples.

Bioremoval of zinc (II) using the bacterial strain M. morganii ACZ05
One ml of overnight culture was inoculated into 500 ml Erlenmeyer flasks with 300 ml nutrient broth spiked with zinc (II) (250, 500, 750, and 1,000 mg L À1 ) and incubated for four days and later optimized to 1-15 minutes. From this, a clear supernatant was obtained by centrifuging an aliquot (10 ml) of the sample at 2,500 rpm. The clear supernatant was subjected to acid digestion, and the residual concentration of zinc (II) ions were measured.

XRD analysis
The diffraction pattern of the biomass powder samples of untreated control and test samples were analyzed using XRD-6000. The diffracted X-ray intensities were recorded using a copper anode, with a step size of 0.0500, in a range of 10-80°, with an XPERT-PRO diffractometer system.

FTIR analysis
The bacterial cells were harvested before and after the adsorption of Zn (II) ions. The bacterial biomass was centrifuged for ten minutes at 3,000 rpm and washed repetitively with 0.9% saline solution at a pH 6.5 (Deokar et al. 2013). Then, they were prepared by the potassium bromide (KBr) method and their profiles were observed (4,000-400 cm À1 ) on an FTIR spectrometer (Bruker, alpha).
2.6. Scanning electron microscopy (SEM) of M. morganii ACZ05 without metal (control) and M. morganii ACZ05 treated with zinc (II) SEM images were taken to make ultrastructural observations of bacterial cells (Srivastava & Thakur 2007) using Field-Emission Scanning Electron Microscope [Carl Zeiss ultra 55; 5 kV with X-ray spectroscopy (EDX)]. The bacterial biomass was harvested, centrifuged, and fixed by glutaraldehyde (2.5%) following dehydration using ethanol. The mounts were dried using tetramethylsilane and coated with gold before documentation.
2.7. Atomic force microscopy of M. morganii ACZ05 without metal (control) and M. morganii ACZ05 treated with zinc (II) The bacterial biomass, harvested following centrifugation and smeared on glass slides, before being dried at a temperature of 22-28°C. SPM 100 Atomic force microscope was used to view and document the samples.

Sorption equilibrium studies
One ml of an overnight culture of Morganella morganii ACZ05 was inoculated into 300 ml nutrient broth spiked with zinc (II) at concentrations of 250, 500, 750, and 1,000 mg L À1 , and incubated. 10 ml of the sample was centrifuged at 2,500 rpm every 24 hours for 4 days. The clear supernatant was subjected to acid digestion and used for AAS analysis (Devika et al. 2014).
AAS Data evaluation for Zn(II) sorption: In the batch sorption studies carried out for the biosorption of Zn(II) ions by Morganella morganii ACZ05 bacterial biomass, the maximum amount of metal ion adsorbed by the bacterial biomass is calculated as follows (Hussain et al. 2004).
Considering the equilibrium, where Zn(II) initial ¼ 250, 500, 750, and 1,000 mg L À1 batches, formulated as C i and Zn(II) final are free unsorbed metal ions determined every 24 hours for four days in each of the above batches. The value of Zn(II) final is determined from AAS and formulated as C e in the following equation.
where Q e is the amount of Zn(II)(mg) adsorbed by per gram of Morganella morganii ACZ05 (mg/g), V is the volume of bulk liquid sample (ml), C i is the initial concentration of the Zn in the (II) solution (mg/L), C e is the final concentration of the Zn(II) in the solution at equilibrium (mg/L) and m is the amount of the Morganella morganii ACZ05 added to bulk liquid (mg). The obtained results are analyzed using the Langmuir adsorption isotherm and the Freundlich isotherm models. The Langmuir adsorption model is given by where Q max is the maximum amount of metal adsorbed under the experimental conditions employed, b is a constant that represents the strength of the adsorption between the biosorbent and sorbate. The linearized form of the Langmuir model states is given by C e =q e ¼ 1=bQ max þ C e =Q max . A plot of C e =q e vs C e will be linear with a slope 1=Q max and y-intercept 1=bQ max . R L is the dimensionless separation factor that is considered in the Langmuir model. It determines the capacity of adsorption between the biosorbate and the biosorbent in the two different phases such as solid and liquid. (Horsfall Jnr & Spiff 2005). It is given by the expression, R L ¼ 1=1 þ bC O where b is the Langmuir constant (mg/L) and C 0 is an initial concentration of biosorbate (metal ions) in the solution (mg/L).
The factor R L determines the direction and feasibility of the reaction between the metal ion and biosorbent, as the isotherm to be either unfavorable ( The Freundlich adsorption model is given by q e ¼ kC 1=n e , where k and n are Freundlich constants, that correlate to the maximum adsorption capacity and adsorption intensity, respectively. Linearized form of Freundlich equation is given by, lnq e ¼ lnk þ 1=nlnC e Sips (1948) proposed an empirical isotherm equation, also identified as Langmuir-Freundlich isotherm, which is often expressed as: where K S is the Sips constant (mg/L), a S is the affinity coefficient (mg/L) and ns is the heterogeneity coefficient.

Kinetic studies
Kinetic experiments were conducted by mixing 10 mg biosorbent of M. morganii ACZ05 with series of 100 ml nutrient broth containing 500 mg L À1 of zinc(II) and incubated at 30°C and optimum pH 6. Samples were drawn every 1, 3, 5,10-and 15-minutes intervals. A mixing speed of 200 rpm was used for this set of experiments. After centrifugation, the supernatant subjected to acid digestion was examined for remaining metal ions concentration (Oh et al. 2009). Lagergren's pseudofirst-order and pseudo-second-order equations were used to model the kinetics of heavy metals.

Fourier transform infrared (FTIR) studies
FTIR analysis of zinc (II) biosorption by M. morganii ACZ05 showed a change in frequency from 3,417.09 to 3,314.04. Figure 1(b) depicts the FTIR spectra of control and zinc-loaded M. morganii ACZ05. The adsorption band shifted from 1,083.14 to 1,079.58.

Scanning electron microscopy (SEM) and energy dispersive X-ray (EDX) studies
SEM and EDX studies were employed to check for the presence and the characterization of metal ions adsorbed to the biomass. Before zinc (II) adsorption, the structure of the bacterial biomass appeared as long, thin, rods. Following the adsorption of zinc, a considerable change in the morphology was detected ( Figure 3). The morphological changes of the bacterial cells which looked fat, spongy, and plumped could be accredited to the treatment of the cell surface with zinc ions. This observation is in agreement with the elemental analysis of cells.
The biomass' energy dispersive spectrum after zinc (II) uptake is revealed in Figure 2. It indicated the presence of separate peaks for zinc, potassium, calcium but no peak for sodium and silicon. The atomic percentage of Zn (at%) was 0.27, while that of Al decreased, from 3.27 to 1.95, while that of C, Mg, P, S, and Cl increased from 30.20 to 30.40, 0.03 to 0.09, 0.15 to 0.59, 0.04 to 0.21 and 0.04 to 0.13 after zinc uptake. The observations indicated that zinc can form covalent bonds with C-, O-and P-containing functional groups during adsorption.

Atomic force microscopy (AFM) studies
AFM was employed to study the cell topological characterization of M. morganii ACZ05 biomass before and subsequently after zinc (II) adsorption ( Figure 3). Observation from the Atomic force microscopy study depicted the collection of zinc precipitates on the external surface of the bacteria. The surface irregularity of bacteria was increased when the bacteria interacted with zinc. The control biomass without zinc exposure had smooth surfaces. Following the sorption of zinc, the cell's morphology was altered into a spindle-like structure.

Adsorption isotherms
The sorption capacity of M. morganii ACZ05 for Zn (II) was determined by fitting experimental data into the commonly used linearized Langmuir isotherm and Freundlich isotherm models.
Application of isotherm models to describe the adsorption of Zn (II) ions by M. morganii ACZ05 with a treatment period of four days was quite unsuccessful, largely due to the low fit of the experimental data on linearization. The Langmuir constant Q max and b values were higher in the second and fourth days. Though the correlation coefficient for Langmuir isotherm model was the maximum on day one of adsorption (R 2 value 0.33) the values were far below the threshold of R 2 . 0.9 (Figure 4(a)-4(d)). But the Freundlich isotherm model was the maximum on day one of adsorption 0.99. The R L value of Zn (II) adsorptions was near 1 for all four days of the treatment period but only day one registered a lower R L value of 0.56. The Freundlich constant n was high on fourth day (n ¼ 4.02) and it decreased continuously for third (n ¼ 1.24), second (n ¼ 0.84) and first day (n ¼ 0.12) of treatment period. From the table above, based on the R 2 values it is quite easily visible that Langmuir, Freundlich and Sips isotherms are not a good fit for the experimental data on all four days (Table 1).

Kinetic studies
The kinetics of biosorption of Zn (II) ions by M. morganii ACZ05, the pseudo-second-order was a perfect fit for the experimental data at the correlation coefficient R 2 ¼ 1, and the equilibrium constant K 2 was 0.0002. While the correlation coefficient derived using Lagregen's equation (pseudo-first-order) was low and insignificant at R 2 ¼ 0.57 and the equilibrium constant K 1 was 0.0007 is shown in Table 2

DISCUSSION
When Morganella morganii ACZ05 was analyzed for its metal removal capacity under varying concentrations of Zn(II) ions, the optimum concentration was detected to be about 500 mg L À1 . Uncorrected Proof X-ray scattering profiles of both the control and test display broad characteristic peaks with weak diffraction patterns. On the other hand, according to Shimazu et al. (2000), sharp and narrow characteristic peaks and diffraction patterns are analogous to the crystalline nature of the sample. This could be ascribed to the polymeric nature of the bacterial biomass. While the bacterial biomass contains cell wall components such as lipopolysaccharides, surface protein, and other biomolecules, it is presumed that the test sample contains extracellular polysaccharides. Heavy metal resistant strains of bacteria are often found to produce extracellular polymeric substances encased within biofilm that protect the cells from metal toxicity and stress (Teitzel & Persek 2003). Both cell membrane biomolecules and exopolysaccharides are materials that are highly polymeric, largely amorphous, and carbonaceous in structure. Since amorphous samples often generate large amounts of background noise in the form of broad peaks, the nature of the X-ray scattering profile is often difficult to interpret. Chemical and structural changes in crystalline compounds are often studied and interpreted together via FTIR along with XRD. Therefore, further studies on the mechanism of biosorption by M. morganii ACZ05 included analysis using FTIR, followed by microscopic techniques.
FTIR analysis was used to check the nature of functional/chemical groups responsible for the biosorption of heavy metals. The characters of the binding sites and their participation in biosorption can be roughly assessed with FTIR. Pagnanelli et al. (2000) reported that the control biomass recorded a spectrum between 3,500 and 3,200 cm À1 region caused by the stretching of the N-H bond in the amino groups. They also reported that the peak similar to the above studies, which is observed  between 3,200 and 3,600 cm À1 might be caused by the hydroxyl group. Kazy et al. (2009) reported that binding of Uranium and Thorium exhibited a strong band at 1,000 and 1,100 cm À1 showing that the carboxyl groups are involved in biosorption. Kazy et al. (2006) observed varying arrangement and intensity of peak at the region between 800 and 400 cm À1 which might be due to the development of strong δ (M-O) and δ (O-M-O) bonds (where M refers to the metal ion).
Minor changes in the peak frequencies between control and test may determine the weakening or strengthening of bond orders in the functional groups involved. A shift in energy frequency of the signature peaks to lower bond orders indicates the weakening of metal-binding sites. Moreover, such interactions of metal ions with the biomass including breaking and formations of new bonds can be retrieved from the increase or decrease in the bond orders observed in the FTIR spectrum of test and control.
In our study, the shift in corresponding peaks was observed in the broad stretch between the wavelength of 3,000 cm À1 -3,500 cm À1 during zinc(II) adsorption, which suggests the presence of a large number of hydroxyl groups or water molecules. The range is also implicated for the stretching vibration of N-H groups present in proteins and their conjugates in the bacterial biomass. Similarly, doublet peaks observed in the region around 2,900 cm À1 may define the C-H stretch representative of the backbone of many major organic compounds including proteins and carbohydrates. The existence of groups like phosphates (P ¼ O and P-O seen in the C-PO 3 moiety) and C(¼O)-OÀ and C-N stretching vibration were also detected as moderate bands around 1,250 cm À1 and 1,080 cm À1 . According to Bremer & Geesey (1991), a broad stretch around 1,000-1,200 cm À1 is caused by carbohydrates (C-O-C, C-O). In particular, the peaks between 1,000 and 1,125 cm À1 determine the existence of uronic acid, o-acetyl ester linkage bonds. Changes in frequencies of the peaks below 600 cm À1 are often attributed to the creation of metal-ligand bond order (Kazy et al. 2006).
These observations overlap for both the test and control samples. The absence of any visible characteristic peaks in the spectrum points out the lack of chemical reaction of the metal ion with metal binding sites if any at the cell surface. Rather, it reveals the possibility of physisorption (or physical adsorption) as the main means of biosorption by M. morganii ACZ05.
The carbonyl, amine, phosphonate, sulfonate, and hydroxyl moieties are the most important groups accountable for the adherence of metal ions on bacterial cell walls. Morganella morganii is a gram-negative bacterium with a cell wall that contains a thin peptidoglycan layer with 20% periplasmic space, and lipopolysaccharides along with specialized transporter proteins inside and outside the lipid membranes (Kumar 2018). Similarly, bacteria in the test samples are presumed to have secreted extracellular polymeric substances as a response to the stress caused by the presence of heavy metals in the media. The extracellular substances of microbes are also a intricate mixture of biopolymers containing polysaccharides, and surface polypeptides, uronic acids, and humic substances. These carry similar functional groups similar to that of the Gram-negative cell membrane. Similarities between functional groups of the cell membrane and EPS could also have resulted in the minor changes in the peak frequencies observed in the FTIR spectrum.
Both XRD and FTIR analysis observations strongly implicate the potential of physical adsorption as a mechanism of heavy metal resistance rather than other means of chemical sorption methods i.e., ion exchange, chelation, complexation. The absence of any chemical alterations to the cell membrane prompts the study of cell surface morphology to determine the extent of localized attachment of heavy metal ions to the surface. The surface morphology of the bacterial biomass was intensively investigated in the present work using AFM and SEM-EDX.
SEM has been quite frequently used correlatively with EDX in biosorption investigations. It helps in the analytical determination of the changes in cellular morphology, size, and volume (bacterial or otherwise), along with an elemental characterization of the cellular surface area. The changes in surface morphology were observed in the biomass of M. morganii strain ACZ05 before and after exposure of Zn 2þ ions. Metal-less control showed distinct cell morphology as rod-shaped bacterial cells, with marked boundaries between them. Whereas, the bacterial cells exposed to metal exhibited various degrees of wrinkling and the loss of distinct peripheries between each cell. The formation of connected cell boundaries could be ascribed to the larger surface area provided by the bacteria for better and faster biosorption ( Jin et al. 2017). Similarly,  shrinkage and wrinkling of bacterial cells along with the development of the rough surface area is implemented to prevent heavy metal toxicity. Studies reported that bacteria grow in size up to a threshold concentration of heavy metals, while excess metal loading causes cell shrinkage and wrinkling. It is argued that live, metabolically active cells have a growth rate that is proportional to their surface area. Stress-induced morphological changes such as decreasing cell surface area available for heavy metal adsorption could be undertaken to increase metabolic activity and survival rate of the bacteria (Naik et al. 2012). Another proposition which follows heavy metal resistance mechanisms in bacteria is the secretion of extracellular polymeric substances (EPS). Metaphorically, biofilms are called 'the city of microbes', and EPS is 'the house of biofilm cells' (Watnick & Kolter 2000;Flemming et al. 2007). EPS affects the bacterial microenvironment by changes in factors such as porosity, density, water content, charge, sorption properties, hydrophobicity, and mechanical stability (Flemming & Wingender 2003). In the present investigation, one of the stark changes between the control and test samples is the amorphous nature of the test sample. XRD and FTIR analysis carry testimony to the same. The metal-loaded bacterial biomass shows vigorous cell-to-cell adhesion and aggregation. It is visible that the number of bacterial cells observed in a unit area has also increased substantially from control to test. Formation of biofilm results in the multiplication of cells inside the microenvironment, leading to water retention, and an increase in adsorption properties. The EPS form a thick matrix inside which metal ions are often sequestered. The dark porous gaps observed in aggregated cell matrices seen in test bacterial biomass are inferred as water channels usually seen in biofilms. The absence of the structures in control is a conclusive indication of the formation of biofilm and secretion of EPS (Asahi et al. 2015).
Moreover, the SEM analysis of metal exposed bacterial biomass showed characteristic bright contrasted sections and new, glossy large particles on the cell surface in comparison to unexposed bacterial biomass. From literature, it is often inferred that localized cation deposition as a granular layer in Gram-negative cell's envelope is usually attributed to the anionic nature of phosphoryl and carboxyl groups present in the cell's envelope. Previously, the presence of such relevant functional groups has been identified in both the test and control samples through FTIR. It follows through for Zn 2þ ions which are hard and bivalent cations (Sar et al. 2001).
In correlation with the results of SEM analysis, the very presence of a specific elemental peak of Zn(II) in the metal-loaded bacterial biomass and vice versa, fetches a conclusive identification of the physical adsorption of Zn 2þ ions by this bacterial strain. Furthermore, changes in the elemental ratio of others are suggestive of different mechanisms behind the biosorption. EDX results also suggest peripheral localization of Zn 2þ ions around the cell wall region, but rather due to ion exchange than sorption to available free sites. Similarly, an increase in the wt % of sulfur from 0.18 to 0.44 is also telling evidence of the adsorption of the ZnSO 4 complexes on the bacterial biomass.
The outcome of the EDX analysis also exposes the rise in the concentration of Mg, P, and Cl ions and cellular release of several new ions such as K, and Ca. Particularly, from these results it can be observed that the wt % of Na has completely disappeared from the cellular biomass from 0.13, while K has newly appeared at 0.22. Concerning metal homeostasis, this is typical of the Na þ /K þ /ATPase efflux pump. This particular phenomenon is usually ascribed to bacterial cells attempting to maintain metal ion homeostasis. Literature gives several instances where bacteria develop heavy metal resistance by adopting a variety of homeostatic functions and possessing different mechanisms for the transport of metal ions. So far, more than 100 P-type ATPases that act as metal transporter proteins have been discovered in bacteria as well as in other organisms (Agranoff & Krishna 1998). The binding of an extracellular counter-ion induces metabolism coupled attachment and blocking of one or more intracellular cations. Among other strategies to reduce toxicity from metal ions, bacteria are also known to possess transmembrane transport proteins (metal ion/proton antiporters) in which proton gradients are utilized to transport metal ions with the expenditure of energy (Agranoff & Krishna 1998).
Yet another distinct feature is the considerable decrease in weight % of aluminum and silica ions from the cellular biomass. Previous studies suggest the exchange of zinc ions in the respective reactive sites in the metalloproteins found in the cell envelope and plasma membrane. The occurrence of the release of new metal ions through metal transporter proteins in the cell membrane can be inferred from the distinct peaks of aluminum and silica (Hase & Finkelstein 1993).
The scrutiny of AFM results is consistent with the findings of the SEM analysis. The control cells that have not been exposed to Zn 2þ ions are found as elongated large in comparison to the metal exposed cells. The metal-loaded biomass appeared much smaller in size as well as exhibited aggregation into clusters. This is in relative correlation to SEM images which showed the lack of distinct boundaries between individual cells. This supports earlier findings of cell shrinkage and wrinkling of bacterial biomass in response to excess metal loading than that can be accommodated on the cell surface as well as intracellular homogenous distribution inside the cell periphery ( Jin et al. 2017). As perceived from the topography scan of the AFM micrograph, the mean depth of the cell was found to reduce by almost fifty percent from control to test sample. It implies the consequent reduction in cell volume after metal ion adsorption. Moreover, the number of bacterial cells in the unit area was observed to have increased, which could indirectly be related to cell aggregation that occurs during biofilm production.
Previous studies argue the presence of surface proteins in the S-layer of the bacterium or the functional groups with strong metal binding tendencies found in the lipopolysaccharides of the Gram-negative bacterial strains which promote the formation of sequestration of the metal ions into biofilm matrix via extracellular polymeric substances (EPS) beyond threshold concentration. Similarly, the absence of integrity and consequent ambiguity in the cell boundaries, and aggregation of bacterial biomass to maintain the reduction in individual cell surface area suggest that these morphological changes have been adapted for stress-induced survival mechanisms (Kazy et al. 2008).
The results concur with each other in the possibility of reduction in cell size and surface area as a metabolism-dependent survival mechanism following physical adsorption, ATPase-mediated transporting of hard cations, and EPS secretion. Similar findings were reported in a study using Pseudomonas and spores of Bacillus species SG-1 strain to remove copper from wastewater (Gavrilescu 2004).
The Langmuir, Freundlich and Sips isotherms are not a good fit for the experimental data on all four days. The only exception to the above data is day one for Freundlich isotherm, wherein the R 2 value is 0.99, making it a very good fit. However, the correlation coefficient for Langmuir isotherms models was extreme at day one of adsorption (0.33) and the values were far below the threshold of R 2 . 0.9. The adsorption capacity constant (Q max ) and affinity constant (b) for Langmuir were largely intermittent values, and thereby considered negotiable.
The Freundlich constants K f and exhibited an upward curve from day one onwards, reaching maximum on day four. Meanwhile, the Freundlich constants K f and n describing the volume and strength of adsorption were quantifiable values on day one (4.44E17 Â 10 À8 , 0.12). Due to the low fit of the experimental data derived up to a treatment period of four days, the treatment period is recommended to be reduced to minutes. The low R 2 values could also be attributed to an insufficient range of concentration of metals available to the cell surface, or the possibility of saturation until equilibrium long before measurement of data.
Biosorption of zinc (II) by dead Pseudomonas putida biomass showed a perfect fit to the Langmuir isotherm model. In this model, metal ions maintained by the biomass growth up to a constant value, due to the liberal reduction of the free attaching sites on the cell wall. But the Freundlich model did not fit the data with the exclusion of the first data points and hence this model is only appropriate at low zinc (II) metal concentrations (Pardo et al. 2003). Elimination of zinc (II) by papaya wood was reported and the initial concentrations of zinc (II) varied from (5-500 mg/L) for the fixed mass of papaya wood (5 g/L) for the sorbate-sorbent interaction time of 60 min. The biosorption of zinc (II) data fit the Langmuir isotherm model with correlation coefficient of R 2 value 0.99 than the Freundlich adsorption isotherm model (Saeed et al. 2005).
In the bioremoval abilities of Zn (I1) ions in aqueous solution using Aspergillus niger, the utmost uptake abilities were 23.70 mg/g of biomass with preliminary metal ion concentrations of 150 mg/L for Zn (I1) ions. The biosorption of zinc ions perfectly fits the Langmuir model (Liu et al. 2006). For the biosorption of zinc (II) in an aqueous solution by coir, the biosorption isotherm data of Langmuir model was not appropriate for the sorption of Zn (II) ions. But sorption of zinc (II) data fit the Freundlich isotherm model with correlation coefficient of R 2 value, 0.99 (Conrad & Hansen 2007). Joo et al. (2010) stated that the removal of Zn (II) by lyophilized cells of Bacillus cereus AUMC B52 and Pseudomonas aeruginosa ASU 6a at a preliminary concentration of 0-200 mg/L and 30 min interaction time. The sorption data fitted properly with the Langmuir and Freundlich adsorption isotherm equations in the investigated range of metal concentration. The results showed that the biosorption of Zn (II) by P. aeruginosa ASU 6a was greater than using B. cereus AUMC B52. The b values of freeze-dried cells of B. cereus AUMC B52 and P. aeruginosa ASU 6a for Zn (II) were 0.048 and 0.035 mg/L, respectively. Chubar et al. (2008) stated that the maximum biosorption ability by Shewanella putrefaciens was 34 mg/g.
The study of the kinetics of metal uptake in the biosorption process is necessary to understand the rate of the process, the optimum conditions required in a batch process reactor, as well as the nature of the adsorption process. Among various models, Lagregen's pseudo-first and Blanchard's second-order equations are employed widely to determine the operating parameters of the biosorption process. Even though the current understanding of the kinetics of the biosorption process is much less, the differences between the correlation coefficient of both kinetics models are utilized to explain the experimental data.
On a similar note, rate kinetics was calculated at particular time intervals of biosorption of zinc onto. morganii ACZ05biomass and data found were plotted against time to determine a desirable kinetic model. The R 2 values as a coefficient of determination specified that the R 2 values for pseudo-second-order were greater than that of the pseudo-first-order.
The layout of a reactor, reaction parameters, and sorption kinetics are necessary to apply in situ and ex-situ (Azizian 2004). The biosorption mechanism, rate of reaction, and its' pathway are important in effluent treatment (Ho & McKay 1999a, 1999b. At the biosorbent-metal ion interface, the kinetics of the reaction hinges upon metal intake and duration of exposure to bacterial biomass (Ho et al. 2000). Meanwhile, biosorption independent of the cell's metabolism, as is predicted to be a rapid method. Bacterial biomass and metal ion equilibrium are achieved in batch reactor conditions in two stages: first 30 mins to one hour followed by the slow forward reaction within 120-180 minutes (Celaya et al. 2000;Choi & Yun 2004;Lu et al. 2006).
The rate kinetics was calculated at particular time intervals of biosorption of zinc onto M. morganii ACZ05biomass and data was found and plotted against time to determine a desirable kinetic model. The R 2 values as a coefficient of determination specified that the R 2 values for pseudo-second-order were greater than the pseudo-first-order. Pseudo-first-order kinetics assume a bimolecular reaction between the adsorbent and the adsorbate, whereas a biosorption reaction between the cell membrane, efflux pumps and exopolysaccharides present in the biofilm can be hardly considered as a bimolecular interaction between a single type of binding site and metal ion. A pseudo first order reaction is often hard to characterise such a highly heterogeneous reaction between the bacterial biomass and the metal ion.

CONCLUSION
Characterization of the mechanism of biosorption reveals the possibility of biofilm production by Morganella morganii due to zinc (II) exposure. Applying linear models of isotherm equations was found to be insufficient to describe the interaction of metal ions on the heterogeneous cell surface. The highly significant fit of pseudo-second-order for biosorption experimental data of Zn (II) ions by M. morganii ACZ05 suggests the involvement of cellular mechanisms for the activity of binding sites of metal ions and other chemical sorption processes. Various mechanisms such as metabolism-independent physical adsorption, and extracellular polysaccharide production, and efflux transport of metal ions through pumps such as ATPases, could be adapted by bacteria to maintain electro-neutrality and osmotic balance, which requires further study.