In this study, 1-Ethyl-3-Methyl-Imidazolium Acetate ([EMIM][Ac]) with purity of 95 wt% and alkali (Na₂CO₃) were supplied from Sigma-Aldrich and used without further purification. The chemical structure of the employed IL ([EMIM][Ac]) is shown in Figure 1. The medium oil sample (API˚ 30.25) and brine were obtained from Weyburn reservoir formation. The properties of Weyburn oil and brine are presented in Table 1. The properties of the sand pack are shown in Table 2.

The selected IL ([EMIM][Ac]) for this study was combined at different concentrations with Weyburn brine. The following steps were taken before the flooding process:

1) Prepare the IL mixtures by adding the following amount of ([EMIM][Ac]), (1000 ppm, 3000 ppm, and 5000 ppm), to the brine, the solution is placed on a stirrer (Cole-Parmer Stable Temp Ceramic Stirring Hot Plate) at 120 rpm for 30 - 45 mints.

2) The viscosity of Weyburn oil (15.355 cP), Weyburn brine (1.07 cP), and displacing solutions was measured by using an A Brookfield DV-II viscometer.

3) An Anton Paar DSA 5000 M instrument was used to measure the densities of the aqueous solutions.

4) KRUSS K100 device was used to measure the surface tension of the displacing mixtures using Wilhelmy plate method. Clean the sample vessel and the plate by acetone first and then by clean water before every measurement. Also, the lower edge of the plate is placed straight and parallel to surface of the liquids.

5) The pH of the displacing phases was measured by a Navi pH Meter. An average value was calculated after every third measurement.

A sand pack column with a bulk volume of 235.7 cm³ was packed with dry Ottawa sand to prepare an unconsolidated sand pack. The average size of the sand, 40 - 80 mesh, was measured using sieving analysis. The PV, porosity, absolute permeability, and fluid saturations were measured and listed in Table 2.

A vertically oriented core holder, with a length of 18.75 cm and an inside diameter of 4 cm, was packed with dry unconsolidated sand. After packing the sand and fixing the caps, the sand pack sample was 100 % vacuumed using a pump until no air bubble came out and then saturated by Weyburn brine to obtain the porosity from the difference between the dry and saturated weight divided by brine density and bulk volume. The column was injected by brine at different injection rates to determine the absolute permeability using Darcy’s Law. After taking the petrophysical properties, the core holder was connected to the core flooding system, a conventional core flooding system, vertically and then the sand pack samples was flooded by the medium oil at a rate of 1 cc/min until no water drop came out from core holder outlet. Total displaced brine represents the original oil in place while the remaining represents the irreducible water saturation. In the next stage, the injection rate was held at a constant rate of 2 cc/min, and the core was placed horizontally and flooded by brine and IL mixtures at different scenarios to obtain the optimum concentration, slug size as well as initiation time. Eventually, all the above procedures were repeated in each experiment with fresh sand to maintain the same properties.

Surface tension (SFT) was considered to determine the Critical Micelle Concentration (CMC) of the displacing solutions at room conditions. The CMC is the concentration at which surfactant solutions surface tension could not be reduced further, as the concentration increased [18] . Moreover, ([EMIM][Ac]) has the capability to reduce the SFT regardless of solutions salinity [16] . As shown in Figure 2, a noticeable reduction in the SFT was observed, as the concentration of the IL increased from 0 ppm to 1000 ppm. Any concentration greater than 1000 ppm resulted in a slight increase in SFT values. It was observed that 1000 ppm of ([EMIM][Ac]) has the highest potential to reduce the SFT of the aqueous solution. So the concentration of 1000 ppm was considered as the CMC of ([EMIM][Ac]). For alkali (Na₃CO₂), as shown in Figure 3, 5000 ppm was considered as the CMC.

The pH values of ([EMIM][Ac]) and Weyburn brine mixtures were measured at 21.5˚C. It was found, in another study, that the pH values of the solutions increased with increasing ([EMIM][Ac]) concentration in the mixtures while the effect of temperature was marginal [19] . As expected, pH values decreased with increasing water content. Additionally, the alkali has the same effect on the pH as presented in Table 3.

Electrical conductivity of displacing phases was measured which increased

with increasing ([EMIM][Ac]) concentration. Similar to the CMC values that obtained from conductivity as function of tributylmethyl phosphonium dodecylsulfate concentration which was in good agreement with the CMC that attained from surface tension measurements [20] . The CMC point obtained from surface tension measurements for ([EMIM][Ac]) + brine solutions was in a great agreement with that obtained from conductivities values versus ([EMIM][Ac]) concentrations.

After preparing the sand pack sample in each experiment, the rock properties were measured, as presented in Table 2. In order to obtain the optimum ([EMIM][Ac]) concentration of displacing phase, three sand pack flooding experiments were performed with three different concentration (1000 ppm, 3000 ppm, and 5000 ppm) and compared with injecting brine alone into sand pack. The flooding process was divided into three stages. First, the sand pack was flooded for one pore volume by formation brine (as secondary recovery mode) followed by one pore volume of IL mixtures. Finally, the sand pack sample was flushed by brine for one pore volume. As shown in Figure 4, the produced oil

was almost identical and the RF values are close to (63% ± 1%) at the end of the first stage. The increase in IL concentration from 1000 ppm to 5000 ppm increased the RF from 77% to 80.64% of original oil in place (OOIP) at end of flooding. Compared to using only water flooding which recovered about 71.17% of the oil in the sand pack sample, the addition of (1000 ppm, 3000 ppm, and 5000 ppm) ([EMIM] [AC]) with the displacing fluid lead to an increase in the oil recovery by 6.57% OOIP, 7.37% OOIP, and 8.77% OOIP, respectively. Finally, as we can see in Figure 4, the increasing on ([EMIM][Ac]) concentration more than 1000 ppm in the mixtures was not efficient to recover more oil economically. Also, the CMC point from SFT and conductivity values were obtained at 1000 ppm ([EMIM][Ac]) + brine solution.

Three experiments were performed to select the optimum slug size (SS). In those experiments, the sand pack samples were initially flooded with 1.25 PV of formation brine, followed by different SS (0.5, 1 and 2 PV) of 1000 ppm ([EMIM] [Ac]), and then the samples were flushed with formation brine. As shown in Figure 5, the total obtained RFs of injecting 0.5, 1, and 2 PV SS were 74.85% OOIP, 77.21% OOIP and 77.23% OOIP, respectively. Regarding the RF results, it is obvious that 1 PV of 1000 ppm ([EMIM] [Ac]) is the optimum SS.

In order to obtain the appropriate initiation time of the chemical slug, the optimum concentration and slug size of ([EMIM][Ac]) were investigated at three different injection times, as shown in Figure 6. First, the sand pack was initially

flooded with 1 PV of 1000 ppm ([EMIM][Ac]), and then it was flushed with 2 PVs of formation brine. The results showed that the RF increased from 71.17% OOIP of only brine flooding to 81.31% OOIP. Second, the injection time investigation occurred when the sand pack was flooded with 0.5 PV formation brine; followed by 1 PV of 1000 ppm ([EMIM][Ac]), then finally, the sample was flushed with 1.5 PV of formation brine, as shown in Figure 5. The accumulative RF for this run was 79.01% OOIP, which is less efficient than that of starting the flooding with the chemical slug. The last examined injection time of the chemical slug started with introducing 1 PV of formation brine to the porous medium, followed by 1 PV of 1000 ppm ([EMIM][Ac]), and then flushed with 1 PV of formation brine. The RF was just 76.95% OOIP, which is less than the final RF of the two previous tests. In conclusion, It is obvious that the earlier the injection of the chemical the better the achieved RF.

In this section, the efficiency of injection 3 PV of 1000 ppm on the RF was studied. As can be depicted from Figure 7 depicts the total RF (84.41% OOIP) and it is noticeably better than that of injecting only Weyburn brine. Apart from the economics, the continuous flooding of the chemical mixture is also more efficient than chemical slug and injection.

One of the upsides of injecting alkali is its ability to react with oil component in order to generate surfactants [21] . Before injection alkali/IL slug, the optimum concentration (5000 ppm) of alkali (Na₂CO₃) based on SFT measurements was obtained. Therefore, 1 PV slug with 5000 ppm Na₂CO₃ was introduced from the beginning of the flooding process and followed by 2 PVs of Weyburn brine to flush the core sample. In comparison to the results of injection brine alone, an extra 4% OOIP was recovered due to the potential of the added alkali. In addition, combing IL with alkali resulted in much better RF than injection just alkali. When 1000 ppm of ([EMIM][Ac]) was added to 5000 ppm (Na₂CO₃) + Weyburn brine slug, the RF was improved from 77.52% OOIP of injecting alkali + Weyburn brine slug alone to 83.95% OOIP, as shown in Figure 8.

To investigate the effect of ILs on wettability alteration, relative permeability (k_ro & k_rw) curves have been calculated and plotted for two flooding experiments (IL + Weyburn brine flooding and Weyburn brine flooding only) under the same conditions. So, k_ro and k_rw of 1000 ppm ([EMIM][Ac]) + Weyburn brine and just Weyburn brine were calculated using step by step graphical technique that was explained by Jones and Roszelle (1978) [22] . As shown in Figure 9, the rock water wetness increases and the residual oil saturation decreases for IL slug flooding comparing with brine flooding. These outcomes confirm the enhancement in the RF results in our study.