Optimization of phosphate oxygen isotope pretreatment measurement method based on phosphate in situ enrichment blanket.

Phosphate oxygen isotope analysis is an effective tool for investigating phosphorus migration and transformation in water bodies. Unfortunately, current pretreatment methods for this technology are significantly limited due to their demanding sample amount requirements, complex operation, and limited scope of application. In order to enhance the efficiency of the pretreatment process, hydrated zirconia was synthesized through liquid-phase precipitation. Zeolite, D001 macroporous resin, activated carbon, and ceramsite were chosen as possible candidate materials for loading purposes. The optimal zirconium loading material was identified through a combination of field enrichment and laboratory elution experiments. The ideal in situ enrichment duration, material dosages, and elution time were ascertained using response surface methodology. The findings showed that D001 resin exhibited superior selective adsorption and elution capacity for phosphate. The response surface optimization yielded the optimal parameters for the in situ phosphate-enrichment blanket: a mass of 13 g for zirconium-loaded D001 resin, an enrichment period of 360 min, and an elution period of 853 min. The attainment of a bright yellow Ag3PO4 solid after purification served as proof of the reliability of the optimization method. The obtained results provide a fundamental basis for the preparation and application of phosphate oxygen isotope analysis in freshwater ecosystem.


Introduction
Phosphorus is an important biogenic element but excess dissolved phosphate in water bodies contributes to eutrophication (Davies et al. 2014). Investigations of the source of phosphorus and cycling process in freshwater bodies are di cult and limited by methodology. Some studies have indirectly investigated the sources, migration, and transformation of phosphorus in water bodies using sediment ngerprinting (Lübeck et al. 2020), microbial community ngerprinting (Heathwaite and Johnes 2015), and spatio-temporal variation in phosphorus speciation (Heathwaite and Johnes 2015). However, challenges remain with quantitative analysis of phosphorus sources and the phosphorus cycle.
Stable isotope analysis is an effective tool for tracing the sources and cycles of elements in nature.
Carbon and nitrogen are widely used in the stable isotope analysis of sources and cycles (Huang et al. 2019). Because there is only one stable isotope of phosphorus, it is impossible to use the stable isotope of phosphorus to trace its sources and cycle ). However, most phosphorus in nature exists in the form of orthophosphate (PO 4 3− ), and oxygen has three stable isotopes ( 16 O, 17 O, and 18 O) (Jaisi and Blake 2014). Therefore, the phosphate oxygen isotope (δ 18 O P ) can be used to trace phosphorus sources and the phosphorus cycle. Research in this area is currently in a stage of rapid development. This technology involves enriching phosphate in environmental samples, separation, and puri cation to obtain a bright yellow Ag 3 PO 4 solid. The phosphorus stable isotope ratio is then measured by high temperature pyrolysis-stable isotope ratio mass spectrometry (Gross et al. 2013). The main issues hindering wide application of this technology are the cumbersome pretreatment method, large sample requirements, and poor method adaptability. Phosphate-selective adsorption materials could be used to achieve in situ phosphate enrichment and highly e cient laboratory elution to optimize the pretreatment method (Tcaci et al. 2019). This will avoid the need for collection of a large number of samples and simplify the subsequent puri cation steps . When optimizing the phosphate oxygen isotope pretreatment method, phosphate-selective adsorption materials could be used to prepare an e cient in situ enrichment device to achieve rapid enrichment and e cient laboratory elution of phosphate.
Hydrated zirconia is a phosphate-selective adsorption material that has promising application prospects (Liu et al. 2018 To address these issues, we used hydrated zirconia as a phosphate-selective adsorption material, and comprehensively investigated the macroporous adsorption, mechanical strength, economics, and availability. We evaluated zeolite, D001 macroporous resin, activated carbon, and ceramsite as potential support materials. The optimum support material was selected after laboratory experiments, and an in situ phosphate-enrichment blanket was prepared using this material. Response surface methodology was used to determine the optimum in situ enrichment time, material dose, and elution time for rapid in situ enrichment and highly e cient elution of phosphate in the pretreatment method. This method could help promote the application of phosphate oxygen isotope analysis in freshwater bodies.

Materials And Methods
Preparation and optimization of the phosphate-enrichment materials Preparation of the phosphate-enrichment materials First, a zirconium solution was prepared by dissolving 150 g of ZrOCl 2 ·8H 2 O in 500 mL of a mixture of 10% HCl, 17% NaCl, 10% ethanol and ultrapure water. Four groups of mixed solutions by this method were prepared. Next, 150 g of support material was added to the solution. The support material was either macroporous D001 resin, activated carbon, natural zeolite, or ceramsite. The solutions were then heated at 60°C for 24 h before ltration. After ltration, the D001 resin, activated carbon, natural zeolite, or ceramsite loaded with zirconium was added to 500 mL of 1 M NaOH. The samples were heated again at 60°C for 24 h. Subsequently, the enrichment material were ltered out and the samples were washed with deionized water until the pH of the eluate was close to neutral. The material was then dried at 60°C for 24 h to obtain zirconium supported on D001 resin, activated carbon, natural zeolite, or ceramsite (Li et al. 2020).

Optimization of the phosphate-enrichment materials
To evaluate the adsorption effect, the Langmuir adsorption isotherm model was used to determine the maximum adsorption capacities of the enrichment materials as follows: where q t is the phosphorus equilibrium adsorption capacity of the phosphate-enrichment material (mg·g −1 ), q max is the maximum adsorption capacity of the phosphate-enrichment material (mg·g −1 ), C is the equilibrium phosphorus concentration in the solution, (mg·L −1 ), and K l is a constant.
A sample (10 g) of each phosphate-enrichment material was weighed into a 100-mL conical ask, and 100 mL of phosphate solution (10, 30, 50, 100, 150, 200, 250, or 300 mg·L −1 ) was added. After shaking at 25°C for 24 h (150r/min), the concentration of phosphorus (C in Eq. 1) in the supernatant was determined by the molybdenum-antimony anti-colorimetric method. The value of q t was calculated from the difference between the mass of added phosphorus and the mass of residual phosphorus in the solution.
The values of C and q t were input into Eq. 1 to obtain K l and q max by tting.
To evaluate the elution effect, the phosphate-enrichment material obtained after the adsorption effect experiments was added to 100 mL of 1 M NaOH. The solution was shaken at 25°C for 24 h and then ltered to obtain the phosphate-containing eluate. Next, 5 mL of the eluate was added to 5 mL of 1 M HNO 3 , and the molybdenum-antimony anti-colorimetric method was used to determine the mass of phosphate. The phosphate elution rate was calculated using the following equations: where m x is the total mass of phosphate removed from the enrichment material (mg); C is the concentration of phosphate in the eluate (mg·mL −1 ), η is the elution rate of the phosphate-enrichment material (%), and m is the mass of phosphate adsorbed on the enrichment material (mg).
The enrichment and elution experiments were carried out in three groups in parallel and the average for the three groups was used for analysis.

Placement and recovery of the in situ enrichment blankets
The in situ phosphate-enrichment blanket was composed of nylon mesh, foam, the enrichment material, and an iron plate (Fig. 1). We determined that the optimum enrichment material was zirconium-loaded resin (Section 1.1), and this was used as the phosphate-enrichment material in the in situ enrichment blanket. To ensure the blanket was easy to use, we gave it dimensions of 5 cm × 14 cm. The blanket was divided every 2 cm so that the material would be in full contact with the water. An iron plate was added to the bottom of the enrichment blanket to weigh it down and ensure it would be completely immersed in the water and not oat on the surface. Foam was placed on top of the enrichment blanket to stop the entire device from sinking into the water. The phosphate-enrichment material was placed in between the foam and the iron plate in ve layers. To prevent weeds on the water surface and waves from affecting the measurements, the enrichment blanket was used to enrich phosphate in water from at least 2 cm below the surface. The nylon mesh allowed for full contact between the enrichment material and the water, and stopped loss of the material into the water. Enrichment of PO 4 3− was conducted in Taiji Lake (Hebei, China) using the in situ phosphate-enrichment blanket. After enrichment, the device was recovered and the support material was removed and placed in a 250-mL Erlenmeyer ask. Then, 200 mL of 1 M NaOH solution was added and the ask was shaken continuously for several hours. Finally, the mixture was ltered through a 0.45-µm lter membrane and the phosphate-containing eluate was retained.
Single-factor and response surface methodology optimization of phosphorus adsorption on the in situ enrichment blanket

Single-factor experiments
In preliminary experiments, we found that the phosphate enrichment effect was poor when the mass of the enrichment material was too low. However, when the mass of the enrichment material was too high, the enrichment effect was not obvious. These results might be related to the speci c hydrodynamic conditions of Taiji Lake. To determine the optimum mass of enrichment material to add to a single enrichment blanket, zirconium-loaded blankets were prepared with different masses (5, 10, 15, 20, 25, and 30 g) of the enrichment material. The blankets were placed in Taiji Lake for 24 h. The optimum mass of the material was determined from the phosphate concentration measured in the eluate obtained after elution with 100 mL of 1 M NaOH solution for 24 h.

Reliability test
There is a lack of isotope fractionation during conventional adsorption and desorption of phosphate (Longinelli et al. 1976; Jaisi and Blake 2010), which means that isotope fractionation will not occur in the present puri cation method  sampling points in Taiji Lake for testing, and evaluated the reliability of the method for generation of bright yellow Ag 3 PO 4 , which was detected by X-ray crystallography after puri cation.

Results
Optimization of the enrichment material Adsorption effect of the phosphate-enrichment material The Langmuir model provided a good t for the adsorption effects of the various materials (Table 1).
Among the four materials, zirconium-loaded resin had the highest adsorption capacity at 19.547 mg·g −1 .
Consequently, this material was selected for phosphate enrichment. Elution of the phosphate-enrichment material Elution was investigated for the four in situ phosphate-enrichment materials (Fig. 3). The elution rates of zirconium-loaded activated carbon, zeolite, and ceramsite were more than 100%, which indicated that these three materials contained phosphate. Although the elution rate of zirconium-loaded D001 resin increased slowly with increases in the phosphate concentration, the mass of phosphate eluted was not larger than that adsorbed.
The molybdenum-antimony anti-colorimetric method (Sinopharm Group Chemical Reagent Co., Ltd., 2007) was used to determine the phosphate mass to establish whether the enriched materials themselves contained phosphate. The phosphate elution rate was calculated using Eqs. 2 and 3. The results for phosphate elution from the four enrichment materials are shown in Figure 4. Phosphate was detected in the eluate for activated carbon, zeolite, and ceramsite loaded with zirconium, which indicated that these three materials themselves contained some phosphate. Mixing of phosphate from the enrichment material with that in the water will affect the oxygen isotope analysis results. Considering these results in combination with those for the adsorption effect, we decided that zirconium-loaded resin was the optimum material for use in the in situ phosphate-enrichment blanket.
Single-factor experiments Figure 5(a) shows the effect of the mass of zirconium-loaded resin on the mass of phosphate eluted. The mass of phosphate eluted was 0.127 mg when the zirconium-loaded resin dose was 5 g, and 0.196 mg when the zirconium-loaded resin dose was more than 10 g. With increases in the mass of zirconiumloaded resin, the number of adsorption active sites will increase. However, too much material will slow phosphorus exchange between the inner material and the water body. This phenomenon will be more obvious in still or slow water bodies. The optimum dose of zirconium-loaded resin was set at 10 g on the basis of the experimental results and in consideration of the economics.

Response surface results
Follow-up experiments were performed using the 17 orthogonal experimental groups generated by Design Expert. Phosphate elution was compared for the experimental groups (  To understand the interaction of each factor and its in uence on the response value, a contour map and three-dimensional response surface map were constructed using the regression equation (Juwar and Rathod 2021). The shape of the contour map can re ect the interaction of two factors. If the contour line is close to elliptical, the interaction of the two factors is stronger, whereas if the contour line is close to circular, the interaction of the two factors is weaker (Jensen 2017). The interactions of three factors in the preparation and application of the phosphate in situ enrichment blanket and the effects on the response value are shown in Fig. 6-8.
The slope of the response surface in Figure 6 was low and the contour was circular, which meant that the phosphate elution time and quality of the enrichment material had no effect on each other. The contours in Figures 7 and 8 were close to elliptical, which indicated that the phosphate enrichment time had an obvious interaction with the quality of the phosphate-enrichment material, and the phosphate enrichment time had an obvious interaction with the phosphate elution time. The slope of the phosphate-enrichment material quality was greater than that of the phosphate enrichment time, which indicated that the phosphate-enrichment material quality had a stronger in uence on the phosphate elution volume than the enrichment time. The phosphate enrichment time slope was greater than that of the elution time, which indicated that the phosphate enrichment time had a stronger in uence on the elution volume of phosphate than the elution time. Therefore, the quality of the phosphate-enrichment material has the most obvious in uence on the phosphate elution volume, followed by the phosphate enrichment time, and nally the phosphate elution time.
The condition optimization function of Design Expert 8.0 software was used to optimize the parameters of the phosphate-enrichment material. The optimum mass of the enrichment material was 13.150 g, the optimum enrichment time was 360.000 min, and the optimum elution time was 852.550 min. The predicted mass of phosphate eluted with these parameters was 0.240 mg. To ensure the in situ phosphate-enrichment blanket was easy to prepare, we rounded the parameters to 13 g for the enrichment material mass, 360 min for the enrichment time, and 853 min for the elution time. The average mass of phosphate eluted was 0.215 mg with these parameters. The actual value was close to the predicted value, which showed that the predicted result was credible.

Reliability veri cation
The reliability of the optimized pretreatment method was veri ed using its ability to generate bright yellow Ag 3 PO 4 . An in situ phosphate-enrichment blanket was prepared using 13 g of zirconia-loaded D001 resin.
After enrichment for 360 min, the resin was eluted for 853 min in the laboratory. The eluate was further puri ed by the optimized puri cation process. Finally, a bright yellow solid was obtained. The X-ray crystallography pattern of the bright yellow solid showed it was Ag 3 PO 4 (Fig. 9). Therefore, the method is reliable.

Discussion
Phosphate oxygen isotope analysis is challenging because of large sample requirements, cumbersome pretreatment methods, and poor method adaptability. These issues hinder wider application of this method in lakes and rivers. In addition to conventional adjustment of test parameters, optimization of the reagent dose, the use of selective adsorption materials for in situ phosphate enrichment, laboratory elution, and puri cation have been investigated as promising pretreatment methods. On the basis of the diffusive gradient in thin-lms technique (DGT) principle, some studies have used an internal zirconium oxide selective adsorption membrane (DGT binding phase) as the enrichment material, and used in situ enrichment, elution, and puri cation steps to optimize the method ). The phosphate oxygen isotope analysis pretreatment method constructed in this paper has the following characteristics: (1) The preparation and use parameters of the in situ phosphate-enrichment blanket are optimized for the target water body, and they could be adjusted for water bodies where the water quality varies. The construction of enrichment blankets with different parameters will allow for accurate control of the speci c material dose, enrichment time, and elution time, which will avoid problems with excessive material consumption, and over lengthy enrichment and elution time.
(2) The in situ phosphate-enrichment blanket is a modular device. Consequently, multiple enrichment blankets could be placed the same point to clarify the results obtained with a single enrichment blanket. The results could be used to improve the method. This could also prevent failure of the experiment caused by operational errors with a single enrichment blanket. (3) The key material in the in situ enrichment blanket is D001 resin loaded with hydrated zirconia. This is combined with a nylon mesh outer layer, which gives the device high mechanical strength. Consequently, the device can be readily transported, placed for sampling, and recycled. Even when crushed, the integrity of the device is maintained. The mechanical strength of the device is higher than that of a conventional zirconium oxide selective adsorption membrane . Therefore, the in situ phosphate-enrichment blanket is bene cial in terms of its adaptability, convenience, and stability.
Currently, the application of phosphate oxygen isotope analysis in water bodies is limited by the complexity of the water matrix. Optimization of the present method provides effective support for the application of phosphate oxygen isotope analysis in freshwater bodies.  (Furer et al. 2006). Phosphate oxygen isotope analysis can be used to explore the relationship between different forms of phosphorus in various environmental media. This method provides a new tool for evaluation of the phosphorus cycle in freshwater bodies and will have great application value.

Conclusions
Liquid-phase precipitation was used to prepare hydrated zirconia. Zeolite, D001 macroporous resin, activated carbon, and ceramsite were evaluated as potential support materials in enrichment and elution experiments. Among the four materials, the D001 macroporous resin was the optimum zirconium support material. Single-factor experiments and response surface methodology were used to obtain the optimum parameters for an in situ enrichment blanket for application in Taiji Lake. These parameters were 13 g of zirconium-loaded D001 macroporous resin, an enrichment time of 360 min, and elution with 1 M NaOH for 853 min The phosphate in situ enrichment blanket performed well on application in Taiji Lake. All experimental groups gave bright yellow Ag 3 PO 4 . The enrichment device could be used to promote the application of phosphate oxygen isotope analysis in freshwater bodies.    XRD diagram of the pretreatment end product