Stone Flour Fertilizers: Using chemical equilibrium modelling to determine quantitative effectiveness of stone flour as a fertilizer

(1)

Stone Flour Fertilizers: Using chemical equilibrium modelling to

determine quantitative effectiveness of stone flour as a fertilizer

MSc Research Thesis Cooper B. Crippen University of Amsterdam

Examiners: Boris Jansen, University of Amsterdam Co-examiner: Chris Slootweg

Date: 07/08/2020

(2)

Main Applicant

Cooper Bliss Crippen Science Park 904 1098 XH Amsterdam +31 20 525 9111 Cooper.crippen@student.uva.nl University of Amsterdam Faculty of Science Masters Chemistry

Science for Energy and Sustainability Student Number: 11985186

Keywords: soil, stone flour, rock dust, rock flour, Visual MINTEQ, theoretical modelling, chemical equilibrium modelling, plant nutrients, bioavailability.

Project Title:

Stone Flour Fertilizers: Using chemical equilibrium modelling to determine quantitative effectiveness of stone flour as a fertilizer

(5)

Abbreviations

Aluminum Al

Calcium Ca

Cation Exchange Capacity CEC

Dissolved Organic Carbon DOC

Dissolved Organic Matter DOM

Institute for Biodiversity and Ecosystem Dynamics IBED Inductively Coupled Plasma Atomic Emission Spectroscopy ICP-OES

Iron Fe

Lithium Li

Magnesium Mg

National Institute of Geophysics and Volcanology INGV

Neodymium Nd Phosphorus P Potassium K Silicon Si Sulfur S Sodium Na Strontium Sr Titanium Ti

(6)

Abstract

Interest in alternative agriculture methods to promote sustainable farming and to replenish increasingly over-farmed land, has turned many back to the use of stone flour, as a means to replenish and rejuvenate over worked or damaged soil systems. While research has been conducted into the effectiveness of stone flour as a fertilizer, there is no quantitative research or data that proves stone flour is effective, and if so, to what degree. This study focuses on the use of Inductively Coupled Plasma Atomic Emission Spectroscopy (ICP), isotopic analyses, and theoretical modelling to gather quantitative data about the release of plant nutrients from stone flour and how effectively plants uptake these essential nutrients. Samples of stone flour, soil, and plant matter were received from Stichting Bargerveen and were to be tested for the elements of strontium, lithium, and neodymium, due to their behavior being similar to essential plant nutrients. However, because of COVID-19 during this project, the ICP and isotopic analysis could not be performed. Therefore, theoretical modelling of the system was performed using Visual MINTEQ to research into the behavior of the system. The model was setup using the composition of the stone flour and soil used in previous field experiments, along with annual rainfall data, temperature, and DOC data from a separate experiment. This chemical equilibrium model was used to look at effects of pH, ionic strengths, and equilibrium rates. The results of the theoretical modelling of this soil and stone flour system showed that pH and ionic strength have a direct influence on the bioavailability of certain plant nutrients. The results also show that changes in pH display notable trends in the speciation of nutrients as well as in the saturation of nutrient minerals.

(7)

1. Introduction

Since the beginning of humankind, agriculture has been a large part of society’s ability to flourish and allow populations to grow everywhere around the world. With this global rise in population, more land is having to be cultivated, along with older lands that have been cultivated for generations (Ramankutty et

al., 2018). This increase in agriculture land use overtime is quite clearly displayed in Figure 2 (Ramankutty et al., 2018). As with everything that is grown within the soil, plants take inorganic nutrients from the soils, so that they can grow. However, nutrients such as nitrogen, phosphorus, potassium, and other macro and micronutrients in the soil that plants need are limited. Thus fertilizers, both organic

and mineral are often added to soils to keep these nutrient levels in balance. But as the necessity for food and other resources grows with the population, more and more lands are becoming less fertile due to over farming and mismanagement of fertilizer use (Söderberg, 2013). The misuse of fertilizers and imbalances can be counter acted though if proper soil treatment is utilized. An example of fertilizer misuse can be the over fertilization of soils with nitrogen, which can lead to a reduction in crop yields and decreased crop quality (Albornoz, 2016). Another example is phosphorus in soils. As phosphorus is a finite resource, preservation and appropriate application is necessary. It has been estimated that roughly 31Mkg of phosphorus accumulation in The Netherlands takes place in agriculture soils, leading to an accumulation of 17kg of phosphorus per hectare of land per year (Smit et al., 2010). In fact, even when working in accordance with EU recommended range of phosphorus application to soils, 70-80% of European soils have an average or high phosphorus status (Römer, 2009). This overapplication of phosphorus within Europe can lead to phosphorus run-off into waters causing eutrophication.

The issue of soil health is not something that is only of importance to humankind, as healthy soils play a role in ecosystems functions, earth’s biosphere, CO2 sequestration, maintains

environmental quality and various other tasks (Doran, 2002). The importance of soil runs so deep that researchers claim it is the difference between survival and extinction of land-based life (Doran, 2002). So, for the sake of humankind and for the health and safety of the environment, the health of soils needs to be maintained. A promising method of conserving the fertility of soils is through the use of stone flour, also known as rock dust or stonemeal, a practice that has been Figure 2 -Estimated total cropland and pasture area from 1700 – 2000 by continent adapted from Ramankutty et al. 2018.

(8)

used for centuries now (Ramos et al., 2014). Stone flour is rock that is ground into a very fine powder. This very fine stone flour is then applied to soils and in studies has shown to positively effect soils through remineralization, rejuvenation of poor soil, balance fertility, conserve natural resources, and promote soil production in a sustainable manner (Ramos et al., 2014 & Leonardos et al., 2000). In fact, all of the 18 elements that are essential for plant growth can be found in naturally occurring rocks and minerals, with the exception of nitrogen (Van Straaten, 2006). This does not mean that all rocks and minerals will contain all the nutrients or high enough quantities of essential nutrients that a plant needs, but rather they are there. However, with all the positives there are also several drawbacks and potential risks to using stone flour. Stone flour is often made from an individual source, meaning the rock only has certain nutrients in a specific ratio. This means for stone flour to meet all the nutritional requirements of plants; several different stone flours must be mixed to obtain all the elements required for proper plant nutrition. Another drawback of stone flour is the potential introduction of heavy metals into soils, which is undesirable if the concentration of these heavy metals is not monitored (Silva et al., 2005). Stone flour also needs to weather in order to release nutrients into the soil and this into the plants. This weathering process varies in length and in some cases may be too slow to properly deliver nutrients.

The use of stone flour also has a financial advantage that artificial fertilizers might not have. Stone flour is cheap to make and is readily available anywhere that has access to nutrient rich rock or stone (Leonardos et al., 2000). A study done in Brazil reveals that the cheap nature of stone flour can be utilized by both modern agriculture and smaller agriculture, as the amount of chemical fertilizer that is needed is greatly reduced (Leonardo et al., 2000). This may change in the future if stone flour use in agriculture is more heavily adopted. Now, stone flour is generally a product that has a limited number of relevant uses. As the benefits of stone flour as a fertilizer becomes more apparent through research and use, this low-priced stone product may soon become a higher priced commodity. Stone flour on its own cannot provide all the necessary nutrients a plant requires in the correct amounts. Stone flour when used in combination with agriculture techniques such as no-till farming, permaculture, manure, compost and other local organic material may be able to improve the performance of stone flour remineralization (Leonardos et al., 2000 & Van Straaten, 2006).

Taking information from available studies of stone flour, they appear to come to the consensus that stone flour does have impact on soil quality, soil nourishment and plant growth (Coventry et al., 2001, Dumitru et al., 1999, Gillman et al., 2001). These studies also focus largely on one of stone flours most beneficial qualities, its’ ability to buffer soil acidity through cation exchange with soils (Coventry et al., 2001). However, largely in part, this is the extent that many of these studies go to and do not delve into mechanistic workings of stone flour or specifics of nutrient movement from stone flour to soil to plant (Coventry et al., 2001 & Gillman et al., 2001). This study plans to be the beginning of unraveling the mechanisms responsible for how stone flour functions. Until this point, little scientific research has been performed in relation to how stone flour works or quantitatively if it affects plant growth and to what extent. The goal of this master’s

(9)

project is to investigate stone flour and to use quantitative measures to provide insight into just how stone flour works and its effectiveness.

2. What is Stone Flour?

Stone flour is often a waste product that is made from mining industries. During the mining process or the processing of a rock for a specific use, there is often large amounts of unused rock or small rocks that are unable to be used for the industries intended purposes. This results in the unused rock being a waste product. This waste rock can then be made into stone flour. An example of this process is the granite mining industry in Spain, where the mining, cutting and polishing of granite slabs to be sold, results in a liquid mixture that contains large amounts of very fine particles of granite (Silva et al., 2005). Once this liquid mixture is dry, a fine granite powder is left, which has the opportunity to be used for agricultural purposes rather than be left as waste (Silva et al., 2005). This is just one of the many ways that stone flour can be made. Large mining quarries also have the opportunity to grind down large pieces of stone to a dust and sell it, rather than going to waste (Silva et al., 2005 & Nunes et al., 2014). This type of operation is also capable of functioning on a much smaller scale as local agriculture can obtain unused rock from various locations. The grinding of these rocks to dust can be beneficial to not only local agriculture, but for personal use as well (Ramos et al., 2014 & Leonardos et al. 2000).

2.1. Stone Flour Nutrients

Depending on the location the stone originates from and the type of rock, the nutrients within the stone or stone flour can vary. However, the idea of removing nutrients from the rock into the soil and plants remains the same despite differences in rock or nutrient content. One of the most common rocks that is used for agriculture application is basalt and this rock has been used since the 1930’s (Gillman et al., 2002). Since basalt is one of the most commonly used rocks, basalt can be used as an example of the nutrients that can be found within these types of nutrient rich rocks.

Chemical Mass Percent (%)

SiO2 49.97 Al2O3 15.99 CaO 9.62 FeO 7.24 MgO 6.84 Fe2O3 3.85 Na2O 2.96 TiO2 1.87 K2O 1.12 P2O5 0.35 MnO 0.20

(10)

Table 1 details the average chemical composition of basalt rock in terms of mass percent (Best, 2002). Element Concentration (ppm) Si 216,00 Fe 105,00 Al 76,000 Ca 65,400 Mg 64,400 Na 20,300 Ti 15,200 K 12,500 P 3030 S 2150

Table 2 - Elemental composition of basalt rock, adapted from Gillman et al. 2002.

Table 2 details the nutrients and their concentrations found in basalt rock used for application onto soils. Plants need a various number of major nutrients that are essential, not only for their survival, but for proper growth. These nutrients are N, K, P, Ca, Mg, and S (Lines-Kelly, 2002). There are also several less essential trace elements required for plant growth such as iron, manganese, copper, zinc, boron, molybdenum (Lines-Kelly, 2002). Table 2 shows that almost all the major elements and trace element, Fe, can be found in basalt stone flour. There are clear benefits to be gained from the use of stone flour in soils, as there are many essential plant nutrients in not only basalt stone flour, but other stone flours as well. The only question is how effective stone flour is at releasing these nutrients so that they be taken up by plants.

Element Contents in rock powder (mmol/kg)

Al 1391 Ca 206 K 503 Mg 195 P 18,9 Mn 5,1

Table 3 – Composition of Lurgi (soilfeed), the stone flour applied in experiments by Stichting Bargerveen, adapted from Weijters et al. 2018.

Table 3 details the contents of the stone flour that was applied to samples that are being analyzed in this experiment (Weijters et al., 2018). This stone flour, known as Lurgi or Soilfeed, was applied to various plots within the National Park Hoge Veluwe by Stichting Bargerveen for a pilot experiment on which this experiment is based. This experiment will be discussed in greater detail in the next section of this research paper. It is important to know the composition of this soilfeed stone flour, because it is the basis of what will be studied in this experiment.

(11)

3. Research Goals

3.1.1. Previous research by B-Ware and Stichting Bargerveen

In field experiments performed by B-Ware and Stichting Bargerveen in the National Park Hoge Veluwe, 10 tons of stone flour (referred to as soilfeed in their experiment) per hectare was applied to several different plots of varying habitat type with varying vegetation type, alongside an identical control group with no stone flour applied (Weijters et al., 2018). This stone flour was applied to each of the plots by hand and was not worked into the soil (Ramos et al., 2014). These plots were maintained and studied for 3 years, with plant samples and soil samples taken annually at a depth between 0 – 10 cm (Ramos et al., 2014). The hypothesis of this study is that stone flour is weathered due to the acidity of the soil, releasing nutrients from the stone flour into the soil, buffering the acidity. In addition, these nutrients released into the soil are taken up by the plants growing in the soil.

3.1.2. Goal of research

The goal of this research project is to create a theoretical model of a soil system with stone flour applied to it. This is being done to estimate or predict the extent of weathering of the stone flour in the various habitat types where vegetation was grown in the Hoge Veluwe. Additionally, this project aims to use the results of this model to quantify the subsequent release of nutrients due to weathering and to estimate the bioavailability of these released nutrients to vegetation from the soil compartment. This will be accomplished by using chemical equilibrium theoretical modelling software (Visual MINTEQ), which is setup using information and data from real stone flour experiment preformed in The Netherlands.

3.2. Knowledge Gaps

While some research exists on the topic of stone flour, many of the papers lack quantitative scientific underpinning of the conclusions that stone flour can effectively be weathered with concurrent release of elements into the soil, where it can be taken up by plants and to what extent this occurs. Therefore, this masters project aims to add information to these knowledge gaps:

• Determine the chemical composition of the soil and stone flour that was used in the field experiment at Hoge Veluwe.

• Use the composition of the soil, stone flour, dissolved organic carbon values, and rainfall data to create a realistic, working theoretical model using Visual MINTEQ.

• Determine if the results from the modelled system are valid, then use changes in pH, ionic strength and temperature to see how these parameters effect the model results.

• Use the results of the various tests and changes in parameters to draw conclusions about the activity and bioavailability of the various nutrients found in stone flour.

(12)

3.3. Research Question and Sub questions

Using the information and data from the previous stone flour experiment preformed at the Hoge Veluwe by Stichting Bargerveen, dissolved organic carbon data from other researchers and rainfall data from nearby weather stations by the Dutch government; can a theoretical chemical equilibrium model be successfully created, and what can the effects of pH and ionic strength have on the bioavailability of nutrients in stone flour?

• According to the theoretical model, what amount of elements released by weathering from stone flour partition into the soil system?

• Can the theoretical model say anything about the bioavailability of nutrients that are present in the stone flour?

o What amounts of certain nutrients are plant available or bound to DOM? o Will the modelled stone flour nutrients dissolve or precipitate out of the

system solution?

3.4. Hypothesis

The hypothesis is that the effects on plant growth and buffering of soil acidity can be linked to stone flour application observed in experiments by B-Ware and Stichting Bargerveen. It is also believed that there is a release of elements from the stone flour that become bioavailable and are partitioning into plants from the soil compartment. This hypothesis is to be tested using theoretical modelling software to closely model the soil system, stone flour and environment of the real experiment, while varying factors such as pH and ionic strength to measure their impact.

4. Methodology

The idea behind this project is to focus primarily on the quantitative analysis of element release from stone flour through weathering and plant uptake of these released nutrients. The samples were going to be tested first with ICP analysis and then isotopic analysis. Due to the scale of this project and the equipment needed, the ICP testing was to be performed at the Institute for Biodiversity and Ecosystem Dynamics in Amsterdam (IBED), in The Netherlands at the University of Amsterdam (UvA). The researcher was then planning to travel to Napoli, Italy to perform the isotopic analysis under the supervision of colleague Ilenia Arienzo of the National Institute of Geophysics and Volcanology (INGV). Both techniques are described in more detail in the next paragraphs. However, due to the time this experiment began and the arrival of COVID-19, it was not feasible to complete the laboratory work presented in this thesis, as university laboratories closed for extended periods of time (see the end of the discussion for more details).

4.1. Samples

The samples consist of plant material from plants that have been grown in soil containing stone flour and without stone flour, as well as soil samples from the soils in question. Thus, allowing the experimenter to perform tests on the stone flour, soil and plant. The samples are relevant to

(13)

this experiment and due to the broad work of Stichting Bargerveen and B-Ware, allows for this experiment to act as a small part of the overarching work being performed by these two companies.

Number Plot Vegetation Type Treatment

1 DH1C dry heathland control

4 DH1S dry heathland limed

7 SZ2C inland dunes control

10 SZ2S inland dunes limed

13 SH1C inland dune heathlands control

16 SH1S inland dune heathlands limed

19 HS2C Nardus grasslands control

22 HS2S Nardus grasslands limed

Table 4 - List of soil samples and vegetation types

Table 4 contains a detailed list of the soil samples that will be tested in this experiment. Soil samples that have not had stone flour applied to them are listed as control, while soils with stone flour applied are listed as limed. Each plot contains 3 control soil samples and 3 limed soil samples, for a total of 6 soil samples per plot. In this set of data, there are 4 different plot types and vegetation types which have been given abbreviations and those are as follows:

• ‘DH’ = dry heathland • ‘SZ’ = inland dunes

• ‘SH’ = inland dune heathlands • ‘HS’ = Nardus grasslands

Most plot types have more than one total plots in which control and limed plants were grown. This ‘duplicate’ plot or ‘plot number’ follows the plot type abbreviation (i.e. DH1, HS2, etc.). For

(14)

this experiment only one of each plot type is required. Thus, plot number chosen is based on which samples from that plot had the most soil and plant material available for analysis. The letter that follows the plot number (C or S) is simply used to differentiate between control samples (C) and limed samples (S). For example, plot SZ2C stands for the second control inland dunes plot.

4.1.1. Sample Location

The plant samples being used in this project were sampled by Stichting Bargerveen in The Netherlands. The samples were taken from a large plot of land with all of the above plot types presented in table 4. Figure 3 details the exact locations of each of the plots. It is important to note that the map also shows other plots that are not being used in this experiment.

4.2. Laboratory Analysis

4.2.1. Inductively Coupled Plasma Atomic Emission Spectroscopy (ICP-OES)

Inductively coupled plasma atomic emission spectroscopy was to be used on the soil and plant samples before performing the isotopic analysis. ICP-OES would allow for determination of major and trace concentrations of elements within samples. Therefore, ICP-OES was to be used

(15)

determine the elemental concentrations of the relevant elements in the stone flour, soil and plant samples. Specifically, to determine the concentrations of strontium, lithium, and neodymium in each respective sample. Measurements of common macro and micronutrients were also be tested for via ICP. ICP-OES was to be performed on a Perkin Elmer Optima8000 ICP machine.

Sample Preparation

All samples that enter the ICP-OES machine must be a liquid. Therefore, the samples were to be extracted in a substantial amount of ultra-pure ammonium acetate. This mixture then would have been centrifuged, and the ammonium acetate would be evaporated out of the solution, leaving a residue. This would have drawn the metal cations out of the sample material. This process was to be repeated at least 3 times to ensure all metal cations have been removed from the sample. After this has been repeated 3 times, the residue would have been dissolved in acid, which was to then be test in the ICP-OES. Ammonium acetate should be used in the metal cation removal from the sample, as this same sample will need to be tested isotopically. Due to the limitations of the isotopic analysis machines, ammonium acetate should be used in place of barium chloride. This sample treatment method is recommended for detecting trace metals, which is applicable in this case with Sr, Li, & Nd. The soil samples were not to be completely destroyed, as the determination of the elements that have moved from the stone flour into the soil is the goal, which is the plant available concentration of relevant elements. Total destruction of soil samples would have resulted in inaccurate data, as the total concentration of the relevant elements in the soil and stone flour would have been the output, rather than just the plant available concentration. For the analysis of Sr, Li and Nd contents of the stone flour and plants, microwave acid digestion was to be performed in the IBED laboratories at the UvA, using an Anton Paar Multiwave microwave acid digestor. Residues of soil samples after ICP-OES would have been destroyed for total element concentration. However, plant samples were to be completely destroyed, as the total concentration of relevant elements throughout the entire sample is the goal.

4.2.2. Isotopic Analysis

Isotopic analysis of the soil and plant samples was to be performed by the researcher at the INGV in Naples, Italy under the guidance of Ilenia Arienzo. This was being done due to equipment constraints on the part of IBED in Amsterdam, as they do not have the equipment necessary to preform isotopic analysis of these samples. The isotopic analysis was to be performed after the OES experiment, as the elemental concentration data of Sr, Li, and Nd gathered from the ICP-OES experiment is relevant to getting accurate results from the isotopic analysis. This isotopic analysis of these samples would have been able to provide insight into the release of nutrients from stone flour and the uptake of these nutrients by plants, through unique isotopic ratios of these elements that exists within the stone flour, soil and plants.

(16)

Sample Preparation

Around 5 grams of all samples (soil, plant, and stone flour) was to be spun in 100ml of 1M ammonium acetate for 12 hours. The resulting solutions then would have been filtered by 0.45-micron filters and then dried on a hot plate. The dried residue is then dissolved in 6N HCl, followed by 2.5N HCl. This 0.5 ml of this acid solution should then loaded onto a quartz column for chromatographic separation of each respective element of interest in each sample (Sr, Li, Nd).

Analytical Conditions

The fractions of each element from each sample were to be dissolved into diluted HNO3 and the

loaded on a degassed zone refined Rhenium filament. For Sr, this should have measured the isotopic ratio of 87_Sr/86_{Sr through thermal ionization mass spectrometry (TIMS) with a}

ThermoFinnigan Triton TI multi-collector mass spectrometer. Sr was to be run with a 87_Sr/86_Sr

isotopic fractionation of 0.1194. The isotopic ratios were to be normalized according to the value of NIST-SRM 987. Similar procedures were to be performed for Li and Nd measurements.

4.2.3. Data Analysis

Using the data gathered from the ICP-OES and Isotopic analysis experiments, there were several things that could be determined by looking at the data. This is the area of the experiment where it can be examined how much stone flour contributes to plant nutrition. When looking at the isotopic ratios results of Sr, this should have given an indication as to the behavior of calcium (Ca), an essential plant nutrient. In this experiment, it was to be assumed that biogeochemically that Sr and Ca behave similarly. The same can be done for isotopic ratios for Li. Which should have indicated the behavior of potassium (K) as, again, the biogeochemical behavior between Li and K is assumed to be similar in this experiment (Kalinowska et al., 2013). The isotopic analysis of Nd would have allowed the researcher to determine the amount of stone flour that was added to the soil and used to grow the plants. This would have been done because the samples are being received from other researchers. The Nd ratios would have allowed for stone flour amounts to be known, as due to previous experiments, the amounts added may be relatively irregular.

4.3. Theoretical Modelling

4.3.1. Chemical Equilibrium Modelling

Along with the ICP and isotopic analysis of the soil and plant samples, theoretical modelling of the system was performed using chemical equilibrium software. Using the samples and the data from previous experiments of Stichting Bargerveen, the soil system with the added stone flour was theoretically modeled using Visual MINTEQ. The details of the Visual MINTEQ are discussed further in an upcoming section. This modelling will give a representation of how the soil system behaves in the model, as well as in a theoretical setting. Ideally, this chemical equilibrium modelling was used to create realistic field conditions that were used to determine the solution concentration of relevant nutrients. These relevant nutrients include Si, Al, Fe, Ca, Mn, Mg, P, Na,

(17)

K, & Ti. This requires the modelling of the system to determine an equilibrium of precipitated phases. Along with this, Visual MINTEQ was used to look at dissolved speciation of the relevant nutrients stated above. The final analysis that was done with Visual MINTEQ was a sensitivity analysis, where the pH and redox potential were varied to evaluate the impact these two parameters have on solution concentration of nutrients, dissolved speciation of nutrients, and nutrient binding to dissolved organic carbon (DOC). Using the information gathered from the models, insight into the bioavailability of the trace elements and nutrients within the stone flour might be made.

Visual MINTEQ Model

Visual MINTEQ is a free chemical equilibrium model that was created, is updated and maintained by Jon Petter Gustafsson since the year 2000 (Gustafsson, 2016). This model is capable of calculating metal speciation, solubility spectra, sorption and pH (Gustafsson, 2016). Visual MINTEQ contains a built-in database of ions, which are used to input the system. Once the components are added, other options such as pH can be specified, or it can be calculated by the system. Visual MINTEQ also presents options to add solid phases, adsorption, surface complexation, gas phases, redox couples and more. The version of Visual MINTEQ that was used in this experiment is Visual MINTEQ 3.1.

Visual MINTEQ Model Parameters

Firstly, the stone flour composition was determined and the amount of stone flour that was spread per 1m2_{was calculated. This was done using data about the stone flour composition from}

the experiment that was done by Stichting Bargerveen in the Hoge Veluwe National Park. In the report by Stichting Bargerveen, the stone flour composition was given in terms of weight percent of metal oxides as determined by XFR. Through calculations, the weight percent of each metal cation was then determined, as seen in table 5. The report by Stichting Bargerveen also states that 15m2_{test plots were used in the experiment and that 225 kg of stone flour was applied to}

these plots. This works out to be 1 kg of stone flour per 1m2_{of land. Now, each cation weight}

percent can be multiplied by 1000g to approximate the mass of each cation that is present per 1m2_{of land due to stone flour application.}

(18)

Oxides from XFR Weight percent (%) Cation weight percent (%) Cation mass of 1000g sample (g) SiO2 47 21.971 219.706225 Al2O3 18.4 9.739 97.38501373 Fe2O3 9.6 6.714 67.14409168 MnO 0.3 0.232 2.323372438 MgO 3.5 2.111 21.10625639 CaO 7.1 5.074 50.74304444 Na2O 5.7 4.229 42.28632648 K2O 5.2 4.317 43.16552017 TiO2 2.2 1.319 13.18551073 P2O5 0.2 0.004 0.436428707

Table 5 – Determination of metal cation mass based on weight percent of oxides and the application of 1kg of stone flour per 1m2_{of land described in Stichting Bargerveen report.}

Visual MINTEQ is a chemical equilibrium modelling system that works mostly in the aqueous state. The main way that stone flour is weathered to release these nutrients is through contact with water. Since there is no information available about the soil pore water content in The Hoge Veluwe National Park, the net precipitation in The Hoge Veluwe National Park area was determined. The net precipitation is the annual rainfall minus the annual evaporation. This number is the amount of rainfall that remains in the soil. The values for rainfall and evaporation were gathered from the Koninklijk Nederlands Meteorologisch Instituut (KNMI). The rainfall and evaporation values were taken specifically from the Gelderland district and the weather station in Deelen. This was done because the Deelen station is in closest proximity to the Hoge Veluwe National Park. The precipitation and evaporation values from the Deelen station were taken and averaged from the past ten years (2009 – 2019), which averaged to a net precipitation of 386mm (Koninklijk Nederlands Meteorologisch Instituut, 2020). Since the model is of a single square meter of land, 1mm of rain is equal to 1L of water per 1m2_{. Thus, in the model it is approximated}

the 1m2_{of land that contains the stone flour received 386L of water in net precipitation annually.}

Using the net precipitation, the mass of each stone flour cation from the 1000g sample was divided by 386L to receive the g/L of each stone flour cation in the 1m2_{plot of land. This value}

was then converted to mg/L and input into Visual MINTEQ.

For an accurate model, the dissolved organic matter (DOM) or dissolved organic carbon (DOC) must be input into Visual MINTEQ. As the presence of DOC will mostly bring negative charges into the soil and this paired with the positive charges of the cation increases the cation exchange capacity (CEC). The experiment performed by Stichting Bargerveen did not measure the levels of DOC is the soils of The Hoge Veluwe National Park. Thus, an estimation of the DOC was obtained from an experiment performed by G.R. Kopittke in The Hoge Veluwe National Park in which they studied effects drought had on soil acidification (Kopittke et el., 2012). The DOC data is not published in the paper, as it became clear drought did not affect DOC levels, but the raw DOC data measured in the experiment was provided by the secondary author of the paper Albert Tietema. The raw data consists of roughly 3000 measurements of DOC concentrations from

(19)

various locations in The Hoge Veluwe National Park from the years 1998 to 2004. All of the DOC concentrations from this raw data were averaged, resulting in an average DOC concentration of 16.7 mg/L. This concentration of DOC was used in the Visual MINTEQ model. Additional component values, not in the report by Stichting Bargerveen, were also added to the Visual MINTEQ model to give a more accurate representation of the soil system. These components are Cl, NO3, & NH4 with concentrations of 6.275 mg/L, 14.633 mg/L, & 0.920 mg/L respectively. These

values were sourced from the same table of raw data that contained the DOC values provided by Albert Tietema.

Visual MINTEQ Tests

The results gathered from the theoretical modeling of the experiment performed by Stichting Bargerveen resulted in several interesting insights into the workings of stone flour. Using the parameters that were described above, several different types of tests were done to determine its influence over the modelled system. First, the types of tests that were done will be discussed, followed by the results of each test on a per nutrient basis.

pH Sweep

The first of the test that was done was a pH sweep. This was done to gather any information on how pH might affect certain aspects of the system, such as freely dissolved nutrients, nutrients bound to dissolved organic matter or changes in speciation. The pH gradient was done between the pH values of 3 – 5, with an increment between each value of 0.1. These values of pH were chosen because the average pH of the soil in the Hoge Veluwe National Park in the Stichting Bargerveen experiment was 4.25 (Weijters, 2018). These values provide a decent range of pH values on both sides of the average and are relatively achievable values in practice. Once the sweep was completed, major nutrients were examined to determine the effect of pH on the binding of these nutrients to dissolved organic carbon. Figure 4 displays what the output of a single pH test looks in the in the Visual MINTEQ interface.

(20)

Ionic Strength

The next modelling test that was done to determine how the ionic strength of the soil solution impacts the activity of the species present in the soil. To begin, the ionic strength is a parameter that details the total concentration of all dissolved ionic charges in an aqueous solution (Blume, 2010). Next, the activity coefficient is a parameter that gives insight into how much a solution differs from an ideal solution (The Editors of Encyclopedia Britannica, 2018). These parameters will be used in several coming figures. The ionic strength that was used for the other modelling tests, such as the pH sweep test, was 0.001. In order to determine the effect of ionic strength on the solution, the modelled system was setup with the initial concentration values of each component as found in table 5, along with the nutrients that were already present in the soil. The pH was fixed at the average pH for the Hoge Veluwe soil found in the experiment by Stichting Bargerveen, 4.25. The system was then run with the ionic strength having values of 0.001, 0.01, 0.025, 0.05, 0.075, 0.1, 0.25, 0.5, 0.75, & 1.0. After each test with a different ionic strength, the activities of all major nutrients were analyzed and plot against the ionic strength gradient to better understand the influence of ionic strength on activity. Figure 5 displays the output from an ionic strength test within the Visual MINTEQ interface.

Speciation

Visual MINTEQ also determines the speciation of each of the nutrients based on the other available elements and compounds that exist in the system. The results of speciation of each of the nutrients took place during the same test for the pH sweep of the system. This means that all of the parameter within Visual MINTEQ were the same for this test as in the pH sweep test described above. Visual MINTEQ returns the activity coefficient of each of the various species that exist at each of the pH values. With these activity coefficients a general idea about how each of the nutrients exists in solution in the soil can be determined. It is important to note that Visual MINTEQ is a theoretical model, meaning that some of the species that are shown to exist in the system might be quite rare in terms of formation in practical application. Some of the species of each nutrient exist in very small quantities, therefore in the figures displaying the speciation, only species that have a large enough activity coefficient are displayed. Figure 5 displays the output from a speciation test within the Visual MINTEQ interface.

(21)

Mineral Saturation

Visual MINTEQ also returns results about the saturation of minerals that are present within the system. Again, these test results came from the same pH sweep test above, thus using the same parameters described above. These results about the saturation of minerals give interesting insight into which nutrients present in the soil system are oversaturated and undersaturated in their various mineral forms. It is important to note since Visual MINTEQ is a theoretical modelling software, some of the minerals that are shown to exist in solution are very rare. This means that some of these minerals might likely will not form in practical application, however, Visual MINTEQ does give insight into how these rare minerals would behave should they form. Figure 6 displays the Visual MINTEQ interface after preforming a saturation test at a single pH value.

(22)

(23)

5. Results

5.1. Nutrient Results

The theoretical modelling tests as described in the Methodology section provided the following results organized per nutrient.

5.1.1. Aluminum Results

Figure 7 displays the ratio of freely dissolved Al compared to the DOM bound Al. The primary y-axis (left) displays the concentration of Al cations in millimolal/L, while the secondary y-y-axis displays the concentration of DOM bound Al cations in millimolal/L, with a pH gradient from 3 – 5 on the x-axis. Analyzing the graph shows that pH has a noticeable effect on the system and displays a clear trend. The clear trend here is that with an increasing pH, the amount of freely dissolved Al cations decreases. While the opposite is the same for the DOM bound Al cations, which increase with increasing pH. Therefore, pH has quite a direct effect on how much Al is available for plants to uptake, as the Al cations that are bound to the DOM are unavailable for plant uptake. Specifically, for Al, it should be noted that in too high quantities, it can be detrimental to plants. Thus, having the highest concentration bioavailable may not necessarily be beneficial. This in short means that as the pH of the soil is increasing, less of the Al will be plant available. As we move onto other nutrients, this will be a trend that continues. It is important to note that while the graph appears to show that the values of freely dissolved Al cations and DOM bound cations are similar, there is quite a large difference in the numbers. In fact, for most of the nutrients the DOM bound cations tend to only account for around 1% of the total concentration of any particular available nutrient.

0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02 3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5 11.755 11.76 11.765 11.77 11.775 11.78 11.785 CO N CE N TR A TIO N DO M A L + 3 ( MIL LIMO LA L/L ) PH CO N CE N TR A TIO N F R EE A L + 3 ( MIL LIMO LA L/L )

RATIO OF FREE & DOM BOUND AL ON PH GRADIENT

Free Al+3 DOM Al+3

Figure 7 - Ratio of the amount of Aluminum cations that are freely dissolved or bound to DOM based on a changing solution pH.

(24)

Figure 8 displays the change in activity coefficient as a function of changing ionic strength. Looking at figure 8 there is a clear influence of ionic strength on the activity coefficient of the Al cations. The activity of the Al cations decreases closer to 0 with an increasing ionic strength. This indicates that as the ionic strength of the solution increases that the Al cations are behaving closer to how they would in an ‘ideal’ solution. In general, based on the information provided by figure 8 that the Al cations in the system prefer when the ionic strength is greater. Figure 9 displays the activity coefficient of the Al cations of the system as a function of pH. Analyzing this figure, we see a similar trend to that of figure 8, with the activity coefficient decreasing with an increasing pH. This again shows a trend of the effect of pH on the activity of the Al cations within the system and shows that the Al cations behave in a more ideal way at a slightly higher pH.

0.00E+00 1.00E-03 2.00E-03 3.00E-03 4.00E-03 5.00E-03 6.00E-03 7.00E-03 0.001 0.01 0.025 0.05 0.075 0.1 0.25 0.5 0.75 1 AC TIVIT Y COE FFIC IE N T IONIC STRENGTH (M)

ACTIVITY OF AL ON IONIC STRENGTH GRADIENT

Al+3 0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5 AC TIVIT Y COE FFIC EIN T PH ACTIVITY OF AL ON PH GRADIENT Al+3

Figure 8 - Measure of activity coefficient of Aluminum cations with varying ionic strength of system.

(25)

Figure 10 displays the species distribution of Al based on a changing pH. This is measured through the activity coefficients of each species of Al from the results provided by Visual MINTEQ for each pH value. It is important to note that figure 10 only includes the major species of Al based on activity coefficient. In total there are roughly 10 different species that exist in the modelled solution according to Visual MINTEQ. However, the majority of these species are in such low concentration, that they are not presented on the figure, as they would not be visible. The only exception to this is the DOM bound Al (/FA2Al+(aq)).

Figure 10 also shows some interesting trends about how the speciation of Al changes with the changing pH. At lower pH’s most of the Al exists as Al cations. As the pH drops, AlH3SiO42+

becomes the most dominant species of Al in solution. There is also an increase of AlOH2+_{as the}

pH increases. It is also interesting to note that the activity coefficient as a whole decreases with increasing pH. This is a trend that coincides with the results of other tests seen in figures 8 and 9. DOM bound Al is included because it is an important indication about the plant availability of Al. However, as discussed before, the amount of DOM bound Al is low in comparison to the total concentration of Al. Because of this, the DOM bound Al is almost not visible on figure 10, however from figure 7 it is known that with increasing pH the DOM bound Al increases in concentration.

0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01 3 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 5 A CT IV IT Y CO EF FCIE N T PH AL SPECIES DISTRIBUTION

Al+3 AlH3SiO4+2 AlOH+2 /FA2Al+(aq)

(26)

Figure 11 displays the various Al-based minerals that Visual MINTEQ determined may exist within the parameters of the modelled system. With all of the Al minerals there is a very clear pattern of saturation as the pH changes. At low pH, almost all of the Al minerals are undersaturated or close to apparent equilibrium. However, as the pH begins to increase there is a clear pattern of oversaturation of almost all the minerals. With the exception of one mineral, at pH of 5 all of the minerals are oversaturated. This clear upward trend in saturation of Al minerals is interesting to note, as it can speak to how the Al within the stone flour will react in practical application. Under these specific conditions, at lower pH the undersaturated minerals will be more likely to be dissolved in solution, while as the pH of the system increases there should be a precipitation of these minerals out of solution.

-10 -5 0 5 10 15 3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5 SAT U RAT ION IN DE X PH

SATURATION INDEX OF AL SPECIES ON PH GRADIENT

Al(OH)3 (am) Al(OH)3 (Soil) Al2O3(s) AlPO4x1.5H2O Halloysite Variscite Diaspore Boehmite Imogolite Kaolinite Gibbsite (C)

Figure 11 - Saturation index of the various aluminum-based minerals that are present in the modelled Visual MINTEQ system.

(27)

5.1.2. Calcium Results

Figure 12 displays the ratio of freely dissolved Ca compared to the DOM bound Ca. The primary y-axis (left) displays the concentration of Ca cations in millimolal/L, while the secondary y-axis displays the concentration of DOM bound Ca cations in millimolal/L, with a pH gradient from 3 – 5 on the x-axis. When analyzing the graph, there is a clear trend that exists that is very similar to that of Al. Just as with Al, here as the pH increases, the amount of freely dissolved Ca cations decreases. While the opposite is true for the DOM bound Ca cations, which increase with increasing pH. Therefore, pH has quite a direct effect on how much Ca is available for plants to uptake, as the Ca cations that are bound to the DOM are unavailable for plant uptake. Just as before, it is important to note that the DOM bound Ca cations only make up a small fraction of the total Ca concentration.

0 0.0005 0.001 0.0015 0.002 0.0025 0.003 3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5 5.2765 5.277 5.2775 5.278 5.2785 5.279 5.2795 5.28 5.2805 CON CEN TRA TIO N D OM CA + 2 (MIL LIM OL A L/L ) PH CO N CE N TR A TIO N F R EE CA + 2 ( MIL LIMO LA L/L )

RATIO OF FREE & DOM BOUND CA ON PH GRADIENT

Free Ca+2 DOM Ca+2

Figure 12 - Ratio of the amount of Calcium cations that are freely dissolved or bound to DOM based on a changing solution pH

(28)

Figure 13 displays the change in activity coefficient as a function of changing ionic strength. Looking at figure 13 there is the same clear trend with Ca the that was seen with Al before. This same trend indicates that as the ionic strength of the solution increases the Ca cations are behaving closer to how they would in an ‘ideal’ solution. With the information provided by figure 13, it shows that the Ca cations in the system prefer when the ionic strength is greater, just as with Al. Figure 14 displays the activity coefficient of the Ca cations of the system as a function of pH. Analyzing this figure, we see a similar trend to that of figure 8 and of the Al cations, with the activity coefficient decreasing with an increasing pH. This again shows a trend of the effect of pH on the activity of the Ca cations within the system and shows that the Ca cations behave in a more ideal way at a slightly higher pH. It is interesting that Al and Ca have a similar trend, as it can give information into how the stone flour behaves as a whole and what environment this stone flour works best in.

0.00E+00 1.00E-03 2.00E-03 3.00E-03 4.00E-03 5.00E-03 0.001 0.01 0.025 0.05 0.075 0.1 0.25 0.5 0.75 1 AC TIVIT Y COE FFIC IE N T IONIC STRENGTH

ACTIVITY OF CA ON IONIC STRENGTH GRADIENT

Ca+2 0.004582 0.0045825 0.004583 0.0045835 0.004584 0.0045845 0.004585 0.0045855 3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5 AC TIVIT Y COE FFIC EIN T PH ACTIVITY OF CA ON PH GRADIENT Ca+2

Figure 14 - Measure of activity coefficient of Calcium cations with varying pH of system. Figure 13 - Measure of activity coefficient of Calcium cations with varying ionic strength of system.

(29)

Figure 15 displays the species distribution of Ca based on a changing pH. This is measured through the activity coefficients of each species of Ca from the results provided by Visual MINTEQ for each pH value, just as with the Al. Figure 15 also only includes the major species of Ca based on activity coefficient. There are several more species of Ca that exist in the system according to Visual MINTEQ. However, the activity coefficients of these species are too low to see on the figure and thus pH induced changes in activity coefficients of these species are also not visible.

When looking at figure 15, there are several insights that we can make. The first is that it behaves in a similar way to Al, in that it as the pH increases the overall activity coefficient of all Ca species decreases. However, it seems that compares to Al, pH does not have as great of an effect on the speciation of Ca in this system. While the majority of the existing Ca species is Ca2+_{, pH does not}

have the same effect as it did with Al. There is a slight decrease in activity coefficient of Ca2+_{, from}

pH 3 to 5, but with Al this change was very drastic. The rest of the species appear to decrease by marginal amounts as the pH increases.

0.004578 0.00458 0.004582 0.004584 0.004586 0.004588 0.00459 0.004592 3 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 5 A CT IV IT Y CO EF FCIE N T PH GRADIENT CA SPECIES DISTRIBUTION

Ca+2 CaNO3+ CaCl+ CaH2PO4+ CaHPO4 (aq) CaOH+ CaNH3+2 Figure 15 - Species distribution Calcium within Visual MINTEQ system based on changing pH.

(30)

Figure 16 displays the saturation index of each of the Ca-based minerals that Visual MINTEQ determined to exist within the modelled system. Just as before with Al, Visual MINTEQ determines all possible minerals based on the contents in the system regardless of rarity. Unlike with Al, all the Ca minerals are undersaturated and do not seem to be greatly influenced by the pH. Most of the increase in saturation for the Ca minerals comes from the very lower pH shift of 3 – 3.3 and then again from around pH 4.6 – 5. This indicates that the pH values between these two ranges are quite stable in terms of the saturation of these minerals. The saturation index for all of the Ca minerals in figure 16 are very low, indicating that the Ca minerals are not coming close to saturating the soil system. This means that the Ca minerals should be able readily dissolve into the system given the proper conditions to do so. This is in contrast with Al, where there was a pH range where Al minerals were undersaturated and a higher pH range where most of the Al minerals were oversaturated. This is another difference that can be draw between the behavior of Ca and Al within this modelled system.

-35 -30 -25 -20 -15 -10 -5 0 3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5 SAT U RAT ION IN DE X PH

SATURATION INDEX OF CA SPECIES ON PH GRADIENT

Ca3(PO4)2 (am1) Ca3(PO4)2 (am2) Ca3(PO4)2 (beta) Ca4H(PO4)3:3H2O(s) CaHPO4(s) CaHPO4:2H2O(s) Hydroxyapatite Portlandite Lime

Figure 16 - Saturation index of the various calcium-based minerals that are present in the modelled Visual MINTEQ system.

(31)

5.1.3. Potassium Results

Figure 17 displays the concentration of the freely dissolved K and the concentration of the DOM bound K within the modelled Visual MINTEQ system. As with the other nutrients, there is clear trend that displays the effect of pH on the concentration of the freely dissolved and the DOM bound forms of K. As with the other nutrients, at a lower pH the freely dissolved K is dominant in the system. This is the first nutrient that Visual MINTEQ has determined at low pH has practically zero concentration in the DOM bound form. The other nutrients began at very low concentrations, but there were not at zero. This is seemingly different here and shows that in this system not until around a pH of 3.3 does some K begin to bind to DOM. However, the rest of the trend continues as the others before, with the amount of DOM bound K increasing with an increasing pH. This trend is also joined by a decrease in the freely dissolved K as the pH increases. This suggests that K general follows the same trend as the other nutrients and increasing the pH of the system causes more K to be bound to DOM.

0 0.00005 0.0001 0.00015 0.0002 0.00025 0.0003 0.00035 3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5 3.36275 3.3628 3.36285 3.3629 3.36295 3.363 3.36305 3.3631 3.36315 3.3632 3.36325 CO N CE N TR A TIO N DO M K+ (MIL LIMO LA L/L ) PH CO N CE N TR A TIO N F R EE K+ ( MIL LIMO LA L/L )

RATIO OF FREE & DOM BOUND K ON PH GRADIENT

Free K+1 DOM K+1

Figure 17 - Ratio of the amount of Potassium cations that are freely dissolved or bound to DOM based on a changing solution pH.

(32)

Figure 18 displays the change in activity coefficient of K with an increasing ionic strength, while figure 19 displays the change in activity coefficient of K with an increasing pH. Figure 18 does slightly differ from the other activity coefficient figures of other nutrients. Figures 8 and 13, of Al and Ca respectively, show a much greater impact of ionic strength on the activity coefficient of the nutrients. In the case of K, this appears to be not a prevalent, as the change in activity coefficient is a very minor one in comparison to Al and Ca. However, figure 19 of the pH appears to follow the same trend as Al and Ca, with a decreasing activity coefficient with an increasing pH. This leads to a possibility that K is not as affected by the ionic strength of the system and will remain relatively the same regardless. However, the activity coefficient of isn’t the best and because it does not shift much upon increasing ionic strength, the only option available to reduce this activity coefficient is through changes in pH.

0.00E+00 5.00E-04 1.00E-03 1.50E-03 2.00E-03 2.50E-03 3.00E-03 3.50E-03 0.001 0.01 0.025 0.05 0.075 0.1 0.25 0.5 0.75 1 AC TIVIT Y COE FFIC IE N T IONIC STRENGTH

ACTIVITY OF K ON IONIC STRENGTH GRADIENT

K+1

Figure 18 - Measure of activity coefficient of Potassium cations with varying ionic strength of system.

0.0032463 0.0032464 0.0032464 0.0032465 0.0032465 0.0032466 0.0032466 0.0032467 3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5 AC TIVIT Y COE FFIC EIN T PH ACTIVITY OF K ON PH GRADIENT K+1

(33)

Figure 20 displays the species distribution of K in the system modelled by Visual MINTEQ based on a changing pH. This is measured through the activity coefficients of each species of K from the results provided by Visual MINTEQ for each pH value, just as with the above nutrients. When looking at figure 20, it is quite clear that K exists in only a few major species with the rest of the species existing in minor species in very small amounts. These minor species exist is small amounts and are not visible on the figure with the major species. Just as with Ca, K seems to take in its major species, then the pH seems to not have an effect. The only noticeable effect that pH has on the major species is that the K+_{species appears to get smaller as the pH increases. The}

other two major species appear to remain the same and there is no large shift between major species such as there was with the Al. Using this information, it is clear to see that both Ca and K have reduced activity coefficient of certain species based on changing pH, but the pH does not particularly influence the speciation of Ca and K.

0.0032458 0.003246 0.0032462 0.0032464 0.0032466 0.0032468 0.003247 0.0032472 0.0032474 0.0032476 3 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 5 A CT IV IT Y CO EF FCIE N T PH GRADIENT K SPECIES DISTRIBUTION

K+1 K2HPO4 (aq) KCl (aq) KH2PO4 (aq) KHPO4- KNO3 (aq) KOH (aq) Figure 20 - Species distribution potassium within Visual MINTEQ system based on changing pH.

(34)

5.1.4. Magnesium Results

Figure 21 displays the concentration of the freely dissolved magnesium and the concentration of the DOM bound Mg within the modelled Visual MINTEQ system. As with the other nutrients, the freely dissolved Mg concentration is located on the left y-axis and the DOM bound Mg is located on the right y-axis. Again, while the concentrations appear to be similar on the figure, the DOM bound concentration only makes up a small portion of the total concentration. As for the contents of figure 21, it shares a similar trend to most of the other nutrients that have been examined thus far. The DOM bound Mg starts at a very low concentration at a low pH, with the freely dissolved Mg starting a higher concentration at low pH. As the pH increases there is a clear decrease in freely dissolved Mg and a clear increase in DOM bound Mg. This is in line with the findings from the other nutrients, which up to this point have all behaved more or less in this same manner. 0 0.0002 0.0004 0.0006 0.0008 0.001 0.0012 0.0014 0.0016 0.0018 3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5 3.298 3.2985 3.299 3.2995 3.3 3.3005 3.301 CO N CE N TR A TIO N DO M MG +2 (MIL LIMO LA L/L ) PH CO N CE N TR A TIO N F R EE MG +2 ( MIL LIMO LA L/L )

RATIO OF FREE & DOM BOUND MG ON PH GRADIENT

Free Mg+2 DOM Mg+2

Figure 21 - Ratio of the amount of Magnesium cations that are freely dissolved or bound to DOM based on a changing solution pH.

(35)

Figure 22 displays the change in activity coefficient of Mg with an increasing ionic strength, while figure 23 displays the change in activity coefficient of Mg with an increasing pH. As with the other nutrients, the trend continues the same. With both an increasing ionic strength and pH, there is a decrease in the activity coefficient of Mg. This is in line with the Al, Ca, K as discussed above.

0.00E+00 5.00E-04 1.00E-03 1.50E-03 2.00E-03 2.50E-03 3.00E-03 3.50E-03 0.001 0.01 0.025 0.05 0.075 0.1 0.25 0.5 0.75 1 AC TIVIT Y COE FFIC IE N T IONIC STRENGTH

ACTIVITY OF MG ON IONIC STRENGTH GRADIENT

Mg+2 0.0028655 0.002866 0.0028665 0.002867 0.0028675 3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5 AC TIVIT Y COE FFIC EIN T PH ACTIVITY OF MG ON PH GRADIENT Mg+2

Figure 22 - Measure of activity coefficient of Magnesium cations with varying ionic strength of system.

(36)

Figure 24 displays the species distribution of Mg in the system modelled by Visual MINTEQ based on a changing pH. This is measured through the activity coefficients of each species of Mg from the results provided by Visual MINTEQ for each pH value, just as with the other nutrients. When looking at figure 24, only two major species exist, with several minor species that are in such small amounts they are not visible in figure 24.The main two species of Mg according to Visual MINTEQ are Mg2+_{and MgCl}+_{, with two minor species. Mg seems to follow the pattern of Ca and}

K, meaning that an increase in pH does not necessarily impact how the Mg will speciate, but rather reduces the activity coefficient through the reduction of a certain nutrient species. This is in contrast with Al, where the increasing pH actively influences which species will be the dominant one, as seen in figure 10. The major species of Mg, Mg2+_{, appears to be the species that}

is mostly affected by the increasing pH. This leads to a reduction in the amount of Mg2+_{until it}

reaches a higher pH of 5, where it is almost equivalent to its other major species MgCl+_.

0.0028645 0.002865 0.0028655 0.002866 0.0028665 0.002867 0.0028675 0.002868 0.0028685 0.002869 0.0028695 3 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 5 A CT IV IT Y CO EF FCIE N T PH GRADIENT MG SPECIES DISTRIBUTION

Mg+2 MgCl+ MgHPO4 (aq) MgOH+

(37)

Figure 25 displays the saturation index of each of the Mg-based minerals that Visual MINTEQ determined to exist within the modelled system. Visual MINTEQ only determines if it possible for a mineral to form and if it determines it is possible, then it is included in the results, regardless of real-life rarity. An example of this can be Magnesioferrite. This mineral is extremely rare, however under these conditions it is possible that it forms, even though in practical application it is highly unlikely. When examining figure 25, it appears to follow a trend of undersaturation, with the exception of Magnesioferrite, a very rare and unlikely mineral as just discussed. This trend is something that was seen with Ca, but not with Al. This means that most of the Mg minerals that have formed in the system can readily be dissolved and the saturation of these minerals will not be a limiting factor in how much of these minerals can dissolve into solution. This effect appears to be influenced by pH. However, even as the saturation increases with an increasing pH, Mg minerals do not come close to the equilibrium line or cross over into the oversaturated area. Thus, the Mg minerals in this modelled system appear to be undersaturated for a quite broad range of pH values.

-30 -25 -20 -15 -10 -5 0 5 10 3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5 SAT U RAT ION IN DE X PH

SATURATION INDEX OF MG SPECIES ON PH GRADIENT

Brucite Chrysotile Mg(OH)2 (active) Mg2(OH)3Cl:4H2O(s) Mg3(PO4)2(s) Sepiolite

Struvite Spinel Magnesioferrite

Figure 25 - Saturation index of the various magnesium-based minerals that are present in the modelled Visual MINTEQ system.

(38)

5.1.5. Manganese Results

Figure 26 displays the concentration of the freely dissolved Mn and the concentration of the DOM bound Mn within the modelled Visual MINTEQ system on a pH gradient. As with the other nutrients, the freely dissolved Mn concentration is located on the left y-axis and the DOM bound Mn is located on the right y-axis. Again, while the concentrations appear to be similar on the figure, the DOM bound concentration only makes up a small portion of the total concentration. This figure of the freely dissolved and DOM bound Mn follows the trend of the other nutrients above. The freely dissolved Mn starts at a significantly higher concentration than the DOM bound Mn. As the pH increases the freely dissolved Mn begins to decrease, while the DOM bound Mn increases. This is a behavior that has been seen with all nutrients thus far and quite clearly shows that the majority of nutrients within the stone flour begin to bind to DOM at higher pH values.

0 0.00001 0.00002 0.00003 0.00004 0.00005 0.00006 0.00007 3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5 0.10945 0.10946 0.10947 0.10948 0.10949 0.1095 0.10951 0.10952 0.10953 0.10954 CO N CE N TR A TIO N DO M MN +2 (MIL LIMO LA L/L ) PH CO N CE N TR A TIO N F R EE MN+ 2 ( MIL LIMO LA L/L )

RATIO OF FREE & DOM BOUND MN ON PH GRADIENT

Free Mn+2 DOM Mn+2

Figure 26 - Ratio of the amount of Manganese cations that are freely dissolved or bound to DOM based on a changing solution pH.

(39)

Figure 27 displays the change in activity coefficient of Mn with an increasing ionic strength, while figure 28 displays the change in activity coefficient of Mn with an increasing pH. The trend with Mn continues as it has before with the other nutrients that have been looked at before. In both figures 27 and 28, an increasing ionic strength and increasing pH lead to a reduced activity coefficient. This means that Mn also behaves in a more ideal manner with a higher pH and a higher ionic strength. This is an interesting point to note, as the other nutrients behave in this same manner as well. This indicates that a large portion of the nutrients within the stone flour react in a similar manner under these specific conditions. This is helpful because having a majority of the stone flour react a certain way makes it an easier material to implement and use to its fullest extent. 0.00E+00 2.00E-05 4.00E-05 6.00E-05 8.00E-05 1.00E-04 0.001 0.01 0.025 0.05 0.075 0.1 0.25 0.5 0.75 1 AC TIVIT Y COE FFIC IE N T IONIC STRENGTH

ACTIVITY OF MN ON IONIC STRENGTH GRADIENT

Mn+2

Figure 27 - Measure of activity coefficient of Manganese cations with varying ionic strength of system.

0.0000951 9.511E-05 9.512E-05 9.513E-05 9.514E-05 9.515E-05 9.516E-05 9.517E-05 9.518E-05 3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5 AC TIVIT Y COE FFIC EIN T PH ACTIVITY OF MN ON PH GRADIENT Mn+2

(40)

Figure 29 displays the species distribution of Mn in the system modelled by Visual MINTEQ based on a changing pH. This is measured through the activity coefficients of each species of Mn from the results provided by Visual MINTEQ for each pH value, just as with the other nutrients. Within this modelled system, there are three major species that Mn primarily exists in. These species are Mn2+_{, MnCl}+_{, Mn(NO}

3)2, which can clearly be seen on the figure above. There are several

minor species that Mn exists as, however the concentrations are so low, that they are not visible on the figure. As with all the other nutrients before, Mn is in its highest concentration at a lower pH and as this pH increases the overall activity coefficient of Mn decreases. The pH only seems to increase the Mn2+_{, as it is the only species that can be seen decreasing as the pH increases.}

This is similar to all other nutrients thus far, excluding Al. This indicates that the pH does not have a large effect on these particular nutrient’s speciation, as generally the speciation stays the same. The pH in these cases only affects the nutrient cation and the rest of the nutrients remain seemingly unchanged. 9.506E-05 9.508E-05 0.0000951 9.512E-05 9.514E-05 9.516E-05 9.518E-05 0.0000952 9.522E-05 9.524E-05 3 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 5 A CT IV IT Y CO EF FCIE N T PH GRADIENT MN SPECIES DISTRIBUTION

Mn+2 MnCl+ Mn(NO3)2 (aq) Mn2OH+3 MnCl2 (aq) MnHPO4 (aq) MnNH3+2 MnNO3+ MnOH+

(41)

Figure 30 displays the saturation index of each of the Mn-based minerals that Visual MINTEQ determined to exist within the modelled system. Visual MINTEQ only determines if it possible for a mineral to form and if it determines it is possible, then it is included in the results, regardless of real-life rarity. Thus, in practical application not all of these minerals will form. Analyzing figure 30 shows that all the Mn minerals are undersaturated within the system parameters. There is a small influence on the saturation due to an increasing pH, however. This increase in saturation is not a major increase and does not bring any of the minerals close to equilibrium or oversaturation. Two of the Mn minerals also appear to not be influence by the pH at all. Based on the results, in practical application it is possible that the Mn minerals and nutrients will be undersaturated or very close to this. This means that the Mn should have a relatively easy time entering solution. -35 -30 -25 -20 -15 -10 -5 0 3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5 SAT U RAT ION IN DE X PH

SATURATION INDEX OF MN SPECIES ON PH GRADIENT

Mn3(PO4)2(s) MnCl2:4H2O(s) MnHPO4(s) Pyrochroite

Figure 30 - Saturation index of the various manganese-based minerals that are present in the modelled Visual MINTEQ system.

Stone Flour Fertilizers: Using chemical equilibrium modelling to determine quantitative effectiveness of stone flour as a fertilizer