Inverse Opal Photonic Crystals as an Optofluidic Platform for Fast

Subscriber access provided by University of Winnipeg Library

Surfaces, Interfaces, and Applications

Inverse opal photonic crystals as an optofluidic platform for fast analysis of hydrocarbon mixtures Qiwei Xu, Seyed Milad Mahpeykar, Ian Burgess, and Xihua Wang ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.8b03179 • Publication Date (Web): 15 May 2018 Downloaded from http://pubs.acs.org on May 18, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Inverse Opal Photonic Crystals as an Optofluidic Platform for Fast Analysis of Hydrocarbon Mixtures Qiwei Xu†, Seyed Milad Mahpeykar†, Ian B. Burgess*,‡ and Xihua Wang*,† †Department of Electrical and Computer Engineering, University of Alberta, Edmonton, T6G 2V4, Canada. ‡Validere Technologies, Toronto, Ontario, M5G 1L7, Canada *[email protected]; [email protected] KEYWORDS: Inverse opal; Photonic crystal; Optical sensing and sensors; Chemical analysis; Hydrocarbon volatility; Detailed hydrocarbon analysis.

ABSTRACT: Most of reported optofluidic devices analyze the liquid by measuring its refractive index. Recently, the wettability of liquid on various substrates has also been used as an important key sensing parameter in optofluidic sensors. However, the above-mentioned techniques face challenges in the analysis of relative concentration of components in an alkane hydrocarbon mixture, since both refractive indices and wettabilities of alkane hydrocarbons are very close. Here we propose to apply volatility of liquid as the key sensing parameter and correlate it to the optical property of liquid inside inverse opal photonic crystals and construct powerful optofluidic sensors for alkane hydrocarbon identification and analysis. We have demonstrated that, via 1 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 24

evaporation of hydrocarbons inside the periodic structure of inverse opal photonic crystals and observation of their reflection spectra, an inverse opal film could be used as a fast-response optofluidic sensor to accurately differentiate pure hydrocarbon liquids and relative concentrations of their binary and ternary mixtures in tens of seconds. In these 3D photonic crystals, pure chemicals with different volatilities would have different evaporation rates and can be easily identified via total drying time. For multicomponent mixtures, the same strategy is applied to determine the relative concentration of each component simply by measuring drying time under different temperatures. Using this optofluidic sensing platform, we have determined the relative concentrations of ternary hydrocarbon mixtures with the difference of only one carbon between alkane hydrocarbons, which is a big step towards detailed hydrocarbon analysis for practical use.

INTRODUCTION The quality analysis of liquids used in many aspects of everyday life is of great importance. Liquids such as wastewater, beverages, chemicals, dairy products and so on must be analyzed from time to time for their own purposes. Specifically, in the oil industry, long term and effective operations of processing plants and the quality of the resulting petroleum products heavily rely on critical analytical measurements performed on incoming and outgoing streams of hydrocarbons. In this regard, detailed hydrocarbon analysis (DHA)1–4 is the main approach for quality analysis since it can provide comprehensive composition data for crude oil feedstocks, fuels, and other petroleum products. Among all the techniques available for DHA analysis, gas chromatography5–9 and fractional distillation10,11 are most commonly adopted. However, a typical analysis in aforementioned technologies can take between half an hour to one hour for a 2 ACS Paragon Plus Environment

Page 3 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


single sample which is mainly attributed to the slow operation process that is performed on a relatively large volume of sample (ranging from a few milliliters up to 100 mL). This is a serious bottleneck for many applications especially in-line analysis. In recent years, optofluidic devices have attracted a great deal of interests for chemical analysis applications12. Most optofluidic devices perform fast chemical analysis of liquids by probing their optical property of refractive index (RI) using transmission and reflection techniques13–15 or surface plasmon resonance16–18. Photonic crystals, such as inverse opal photonic crystals, are promising candidates for optofluidic sensing19–23 because of their tunable optical responses24. However, a low resolution of 0.001 RI unit/nm25 prevents their applications to hydrocarbon analysis, since various hydrocarbon liquids have very close refractive indices (see table S1 in Supporting Information). Recently, the wettability of liquid on various substrate materials22,26,27 have been demonstrated to identify liquids, and show remarkable selectivity based on controlled surface modification of substrate materials using alkylchlorosilane chemistry. In order to improve selectivity while maintain sensitivity in probing wettability, Lee et al. proposed to integrate several different function groups onto the same substrate and successfully distinguished between various structurally similar toxic chemicals28 using principle component analysis. However, some chemicals like various alkane hydrocarbons have the same function group and similar wettabilities. Here we propose to use another important property, volatility of chemicals, for hydrocarbon identification and analysis of each component in a hydrocarbon mixture. Thanks to the possibility to couple strong changes in optical properties (angular and wavelength dependence of scattering) of 3D photonic crystals to liquids filled inside photonic crystals, it was shown that accurate analysis of volatility of different pure alkanes, as well as detection of the relative concentration of components in binary mixtures (composed of two 3 ACS Paragon Plus Environment


Page 4 of 24

components) can be done in an inverse opal photonic crystal29. Despite its simplicity and relatively rapid nature, the approach mentioned above was limited to analysis of pure hydrocarbons or binary mixtures composed of one volatile and one non-volatile component. Moreover, since the test is performed at room temperature, its duration ranges from a few minutes up to half an hour (depending on the evaporation rate of the volatile component at room temperature) which is far less than ideal for in-line analysis applications. In our demonstration, we probe the volatility of various mixtures in various heating conditions, and extend our analysis to the ternary mixture (composed of three volatile components) by leveraging a modified version of the Penman Equation, a classic equation describing evaporation rate for water30, to theoretically calculate evaporation rates for different pure hydrocarbons. By correlating theoretical calculations to experimental data of binary mixtures, we have extracted all fitting parameters in the modified Penman equation and successfully used the equation to determine relative concentrations of ternary mixtures with high precision. More importantly, by performing the test above the room temperature through highly controlled heating, as well as requiring a very small volume of sample (in the order of nanoliters), the analysis speed has been significantly increased and the test duration has been reduced to a few seconds. Finally, this new technique is not limited to the presence of a non-volatile component at the test temperature and is capable of identifying closely related alkane hydrocarbons with minimal carbon content difference in both binary and ternary mixtures. The developed analysis method enables the great potential of 3D photonic crystals as a simple, fast and low-cost optofluidic platform for compositional analysis of hydrocarbon-based mixtures.

RESULTS AND DISCUSSION 4 ACS Paragon Plus Environment

Page 5 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


2.1 Setup and experimental phenomena Three dimensional photonic crystals made of inverse opal films (IOFs), as displayed in Figure 1A, were used as an optofluidic sensing platform for characterization of the liquid drying process inside the IOFs. A photograph of the setup used for data acquisition and control of the drying process is shown in Figure 1B. Normally, once the IOF is fully infiltrated with different volatile chemicals under room temperature, it can take several to tens of minutes before the liquids inside the IOF are completely vaporized. Here, to achieve ultra-fast analysis results delivery, the data acquisition and control system was designed to accelerate the process by providing controlled heating to the sample under test.

Figure 1. (A) A photograph of an inverse opal film. The inset image shows details of the optical response of a partial-filled inverse opal structure inside the film. (B) Experimental setup: heat control governs the temperature of the heater, reflected light is transferred to a spectrometer by a

5 ACS Paragon Plus Environment


Page 6 of 24

fiber, and spectral information is further recorded and analyzed by computer. Camera is used to ensure that each experiment is conducted in the same region on the IOF.

Figure 2. (A) Practical spectrum of a dodecane-undecane mixture (0.75-0.25 volume ratio) at 50oC for stages 1 (fully filled with liquid), 3 (half liquid in the structure), and 5 (empty structure), and their corresponding simulations. (B) Typical time evolution of spectrum of the same mixture; inset plot shows the spectrum when sudden peak change happens at t1=2.0s, due to disappearance of over-layer chemicals. (C) Time evolution of total reflection intensity of the 6 ACS Paragon Plus Environment

Page 7 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


same mixture: drying time is defined as the time span between stage 1 and stage 5. Stage 5 is distinguished based on total reflection intensity while stage 1 is based on sudden peak change. Using the setup, we were able to bring liquid samples to desired temperatures in less than a second and observe and record the evolution of optical response (spectral intensity) of the liquidfilled IOFs as a function of time. Taking a dodecane-undecane mixture (0.75-0.25 volume ratio) at 50oC as an example, the measured spectra of three distinct stages are shown in Figure 2(A), when the IOF is fully filled, half infiltrated and totally empty, along with the simulated spectra using the finite-difference time-domain method (Lumerical FDTD Solutions). Note that although the general shapes (existing 3 peaks) of reflection spectra are similar between simulations and experimental results, the discrepancy of shift trends between simulations and experimental results was observed. Therefore, in the following analysis of mixed hydrocarbon chemicals, the integrated reflection intensity across the entire spectrum (450 nm to 700 nm) was used and the simulation of spectrum was not taken into consideration. However, the evolutions of the spectrum and the spectrum itself may provide rich information for future analysis, including the structure of the photonic crystals, the pore sizes, the type and volatility of liquid and filling percentage in the structure, etc. Therefore, further comprehensive analysis and accurate simulations of the time-varying optical response of the chemical-IOF system should be carried out in the future. A typical sample plot of the obtained time-varying data of the same mixture is displayed in Figure 2B. By further integration of the spectral intensity over the wavelength range, one can obtain the evolution of total reflection intensity as a function of time illustrated in Figure 2C. This is needed for characterization of the drying process, where the evolution of total reflection intensity can be described by five stages29, where stage 1 and 5 represent the onset and completion of the drying process, and stage 3 marks mid-point of the drying process (half 7 ACS Paragon Plus Environment


Page 8 of 24

volume of liquid remaining inside the structure). Note that according to the percolation theory in the previous report29,the degree of disorder, quantified from the Fourier transform of the inverse opals as the portion of scattered intensity outside the normal direction, increases with the filling percentage of the inverse opals as drying progresses and reaches maximum when the structure is halfway filled (stage 3). Consequently, total normal reflectance would reach its minimum at stage 3. Experimentally, stage 1 can be identified by observation of a sudden change in peak positions in the spectral data (marked as t1 in Figure 1B) which is due to disappearance of extra liquid layer sitting on top of the IOF chip (over-layer liquid) during sample loading29; stage 3 is identified as the minimum-intensity point; and the corresponding time for stage 5 is identified as the point after which overall reflection intensity remains fairly stable and peak positions in spectral intensities no longer shift in terms of wavelength, indicating a completely empty structure. Figure S1 in Supporting Information shows the photographs of the drying process, illustrating a brightness (intensity) change in the practical experiment. After identification of the onset and completion moments for the evaporation process, sample drying time can be defined as the time span between stages 1 and 5. The optical measurements of drying times for all samples are conducted via the same method at normal angle. 2.2 Calculation of drying time As drying evolves with time, liquids inside the pores would form different disorder filling patterns, and it can be well described by percolation theory29, which, however, is not able to provide any measure of chemical volatility. To this end, we modify Penman Equation to calculate evaporation rates for the measure of volatility of different chemicals inside an IOF. Penman Equation, an empirical equation but with strong physical basis used for describing


Page 9 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


evaporation from an open water surface, was developed by Howard Penman30 and re-written in metric units by Shuttleworth31, which takes the form: E=

mRH + 6.43γδ ( p ) λ (m + γ )

(1)

where E is the evaporation rate (in mm/day); m is the slope of saturated vapor pressure curve (in kPa/oC); RH is the heat fluxes available for evaporation [in MJ/(mm2day)]; λ is enthalpy of vaporization (in MJ/kg); δp=p*-p0 (in kPa) is the pressure deficit, where p* is the saturated vapor pressure determined by different temperatures and p0 is the actual vapor pressure; γ is the psychrometric constant defined as:

γ=

C p ( air ) P

λM ratio

(2)

where Cp(air) is the heat capacity of air (in kPa/oC), P is the atmospheric pressure (in kPa), and Mratio is the molecular ratio between water and air. Saturated vapor pressure is characterized by another empirical equation, Antoine Equation32:

p * = 10

A−

B C +T

(3)

where T is the temperature, and A, B and C are chemical-specific constants. Notably, Penman Equation is designed exclusively for water (single liquid) in an open area. We extended application of the equation to other pure chemicals and their mixtures by modifying it into:

 mR + 6.43γxδ ( p )   E = a H λ (m + γ )  

(4)

where we believe the added parameter, a, is a both crystal and chemical-related parameter; the additional x is the molar fraction of component in a mixture, and x=1 for single pure liquid. x is introduced in modified version based on the Raoult's Law33, which states that the partial vapor 9 ACS Paragon Plus Environment


Page 10 of 24

pressure of a solvent equals to the saturated vapor pressure of the pure solvent at the same temperature multiplied by its molar fraction, as the pressure deficit for a pure solvent is ‫ ∗݌‬− ‫݌‬଴ while partial pressure deficit ‫ ∗݌(ݔ‬− ‫݌‬଴ ) in mixtures. Then we developed an algorithm, as shown in Figure 3A, that uses evaporation rates obtained from the modified Penman Equation to calculate drying times for both pure liquids and chemical mixtures. In each time step, the evaporation rates for each component are calculated with the updated molar fractions determined from remaining volume of each component; time is added up until no liquid is left in the structure (empty pores), and the resulting time is the drying time for pure liquids or chemical mixtures (for single pure liquid, molar fraction remains 1 throughout the operation). This algorithm, since it is capable to attain drying times from known concentrations, is called “forward algorithm”. To further solidify it, we compare two timespans from stage 1 to stage 3 and stage 3 to stage 5, three distinct stages representing an IOF fully infiltrated with liquids, half volume liquids left in the IOF, and an empty structure, both in simulations and practice. Experimental and simulated results are shown in Figures 2A, B and Figure 3B, respectively (taking dodecane-undecane mixture with 0.75-0.25 volume ratio under 50oC as an example). One can conclude that it takes 7.2 s from drying start (stage 1) to half liquid point (stage 3), and another 7.9 s before a completion of drying (stage 5) in experiment, while 7.15 s and 7.94 s for corresponding time spans in simulated drying process. More importantly, in practical experiment, evaporation rates are not linear, but gradually decrease via time evolution, and this decreasing trend is well captured by our simulation in Figure 3B, where the slop of the red curve decreases when time increases (black straight line as the reference), indicating smaller evaporation rates as drying evolves in time.


Page 11 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3. (A) Flowchart of forward algorithm for calculation of drying times of both pure chemicals and mixtures. (B) Simulated time-evolution (red curve) of left volume in a dodecaneundecane mixture with initial 0.75-0.25 volume ratio. The evaporation rate of the mixture declines with time (see black straight line for reference). Simulated time spans from stage 1 to stage 3 and stage 1 to stage 5 match with that in practice in Figure 2B. 2.3 Liquid identification and concentration analysis Based on the idea that evaporating rates vary with liquids with different volatility, an IOF-based optofluidic system then can be utilized for liquid identification and relative concentration analysis by measuring drying times of liquids inside the structure. To test the validity as well as the accuracy of this method, we chose four hydrocarbons that are closely related, decane (CH3(CH2)8CH3), undecane (CH3(CH2)9CH3), dodecane (CH3(CH2)10CH3) and tridecane (CH3(CH2)11CH3). For simplest case, as shown in Figure 4A, one can easily differentiate four similar pure hydrocarbons with their distinct drying times under different temperatures. Note that 11 ACS Paragon Plus Environment


Page 12 of 24

the drying time for decane under 70oC is not included as decane is extremely volatile and is beyond the capability of the testing system. For the purpose of differentiating relative concentrations of a two-component mixture, the drying times for all the concentration combinations have to be predicted. At first, drying times of the mixture were measured for five concentration combinations (e.g. volume ratios of 1-0, 0.750.25, 0.5-0.5, 0.25-0.75 and 0-1 were chosen for the dodecane-undecane mixture) under three temperatures, 50oC, 60oC and 70oC. Then the evolution of drying times for different concentrations was theoretically predicted by the forward algorithm, with minimal deviation between the simulated evolutions (blue solid curves in Figure 4B) and the five experimental data (red dots in Figure 4B). Fitting parameters from the characterized blue curve are summarized in Table 1. After the drying times for all possible concentration combinations are calculated, the concentration value of any unknown dodecane-undecane mixture can be obtained through its measured drying times under three temperatures. For each drying time measured under certain temperature, only one concentration value can be obtained from the simulated evolution (blue curve in Figure 4B). The final predicted concentration is the average value of three obtained concentrations under 50oC, 60oC and 70oC. To validate the above-mentioned method, we conducted drying time measurements on three samples of dodecane-undecane mixture with randomly selected relative concentrations, and the predicted concentrations show good agreements with selected values (see dashed line in Figure 4B).


Page 13 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4. (A) drying times for different pure chemicals under three temperatures; dry time of decane under 70oC is not included in the figure as it is too volatile to catch the starting point of decane’s drying process before temperature is heat up to 70oC. (B) experimental data and simulated drying times for different two-element mixtures under three temperatures: experiments were carried out on the dodecane-undecane mixture with five different concentration combinations, 0-1, 0.25-0.75, 0.5-0.5, 0.75-0.25 and 1-0, represented by red dots; evolutions of drying times for all possible concentration combinations, shown as blue curves, are characterized with minimal deviation between simulated drying times and practical points. To test the capability of differentiating unknown binary mixtures, drying times of three randomly selected samples (dodecane-undecane), sample 1,2 3, are measured, with drying times of 14.2s, 19.3s, 22.8s under 50oC, 6.0s, 8.0s, 9.2s under 60oC, and 2.7s, 3.3s, 3.6s under 70oC, as indicated by the dash lines in three different colors under each temperature; and their relative concentrations are derived upon the simulated curves; predicted concentration values of the mixtures are 0.28-0.72, 0.56-0.44, and 0.77-0.23, and real concentration values are 0.29-0.71, 0.56-0.44 and 0.78-0.22. Note that measurements are carried out under three temperatures, and the final predicted



Page 14 of 24

concentration value for each unknown sample is averaged from three calculated values under three temperatures. Table 1. Fitting parameters of PE a Hydrocarbons RH

p0

50oC

60oC

70oC

50oC

60oC

70oC

dodecane

0.1

13.5

31.5

74

0.127

0.343

0.430

undecane

0.1

33.8

45.9

127.0

0.310

0.250

0.900

For the prediction of ternary-mixture concentrations, we chose the tridecane-dodecaneundecane mixture as an example to show the procedure of analysis. To this end, evolutions of drying times under different temperatures of all possible two-element mixtures were first characterized (i.e. simulations of drying-time evolutions of tridecane-dodecane, tridecaneundecane, dodecane-undecane mixtures with minimal overall deviation between measured points and simulated curves) before calculations of drying times of three-element mixtures. Results are shown in Figure 5A. With the fitting parameters in supplementary Table S2 derived from the optimal simulations of binary mixtures, we are able to predict the drying times for threecomponent liquid, with any random concentration combinations, via the forward algorithm. The reverse problem of determining relative concentrations of each component in a ternary mixture is not as straightforward as the reverse problem in a binary mixture. Previously, Li et al proposed an algorithm of quantitative principle component analysis to quantify each of the components in a mixture34. Here, we proposed a “backward algorithm”, as displayed in Figure 5B, to determine the relative concentrations after the drying times of an unknown ternary mixture is measured under three temperatures. The algorithm starts with calculating all possible combinations of each components in the ternary mixture which have the drying time at a particular temperature. As the 14 ACS Paragon Plus Environment

Page 15 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


calculated drying times of all combinations in the ternary mixture are done using forward algorithm with the derived fitting parameters listed in Table S2, possible combinations leading to the experimentally measured drying time at different temperatures are shown in Figure 5B (Pink region for 50oC, Green region for 60oC, Blue region for 70oC). A deviation margin value in terms of drying times should be set in the plot (e.g. 0.2s), since the experiments have certain variations in the measurement of drying times. Therefore, all possible combinations will form three regions for three temperatures, indicated by three color “stripes” in the inset plot in Figure 5B. For example, if the margin value is 0.2s, under 60oC a tridecane-dodecane-undecane mixture with many different combinations of relative concentrations would have the predicted drying times which are within the deviation of 0.2s with the measured time. The procedures are the same for the other two temperatures. The overlap between the three color stripes represents all possible concentration combinations under all three temperatures. In practical, only one defined combination should be selected from the overlapped area for the real concentration combination. Hence, we continuously decreased the deviation margin value to reduce the overlapped area of three color strips, until the best estimation of concentration combinations was made when there was no overlap between the three color strips. The inset color plot in Figure 5B provides a simple demonstration of solving the reverse problem for 0.333-0.333-0.333 volume ratios of a tridecane-dodecane-undecane mixture. To validate the backward algorithm, we measured drying times of the three-element mixtures with four different concentration combinations, 0.25-0.250.50, 0.25-0.50-0.25, 0.333-0.333-0.333, and 0.50-0.25-0.25 volume ratios under three different temperatures. Predicted concentrations are shown in a 3D plot in Figure 5C, which shows high accuracy with maximum 0.008 concentration difference in all cases. We also validated our



Page 16 of 24

method using additional two IOF samples, and results show good accuracy (see Figures S3 and S4, and Table S5 in Supporting Information). The current demonstration has limitations when organic compounds have extremely similar volatilities. In the future, the demonstrated method of probing volatility could be improved in the identification and analysis of various volatile hydrocarbon liquids, volatile/non-volatile mixture of organic liquids, and organic/water mixtures using advanced analysis algorithms (e.g. machine learning).

Figure 5. (A) Experimental drying times and simulated ones under three temperatures for different two-element mixtures: experiments were carried out on tridecane-dodecane, dodecane16 ACS Paragon Plus Environment

Page 17 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


undecane and tridecane-undecane mixtures with five different concentration combinations, 0-1, 0.25-0.75, 0.5-0.5, 0.75-0.25 and 1-0; evolutions of drying times are simulated with minimal deviation between practical points. (B) Flowchart of backward algorithm with inset color plot demonstrating the method of deriving concentrations via drying times of three temperatures. (C) Results of concentrations for four tridecane-dodecane-undecane mixtures (0.25-0.25-0.50, 0.250.50-0.25, 0.333-0.333-0.333, and 0.50-0.25-0.25). CONCLUSION Acquisitions of accurate and real-time data of hydrocarbon oil (e.g. compositions and relative concentrations) before and after refinement are always preferred, since one can take timely actions to alter processing conditions, thus reducing quality control risks and enhancing production yields and refiner profitability. According to the experimental results, the limits of detection for binary and ternary hydrocarbon mixtures are 1% and 2.7%, respectively. Further improvement in experimental set-up and prediction algorithm can be carried out to achieve smaller detection limit. In this case, IOF optofluidic sensors, capable of identifying chemicals and accurately differentiating mixture concentrations in tens of seconds, can be fabricated into a low-cost real-time data acquisition device for chemical identification and analysis purposes. More importantly, with careful and precise temperature control, the structure can be readily applied for analyzing mixtures with more components, which shows great potential for ultra-fast DHA analysis in practice. EXPERIMENTAL SECTION Four different hydrocarbons with close refractive indices were utilized, decane, undecane, dodecane and tridecane, to demonstrate the ability of differentiating similar hydrocarbons by probing their volatility. The refractive indices of the four chemicals provided by the 17 ACS Paragon Plus Environment


Page 18 of 24

manufacturer (Sigma-Aldrich) are summarized in Table S1 in Supporting Information. IOFs used in this work are made of silica, fabricated on silicon substrates, and the fabricating process and the surface functioning process are introduced previously29,35. As shown in scanning electron micrograph (SEM) images in Figure S2 in Supporting Information, the large-area crack-free structure has 9 layers with the pore radius ~150nm, thanks to the colloidal co-assembly technique described previously35. The pore size of the IOFs is small so that the measurement only consumes nanoliters of samples, thus speeding up the whole process. Furthermore, in such structure, there would be large change in degrees of disorders within the lattice as components dry up. This will provide us with information in optical properties, enabling us to differentiate closely related hydrocarbons. In terms of the different choices of photonic crystals, the crystalto-crystal variation will change the fitting parameter, but for the same substrate, the experiments were repeatable and consistent results were obtained. To validate the idea, total three IOF chips (a, b and c) were utilized in this work. In the case of binary mixtures, Figure 4(B) shows the practical and predicted dry time on chip b and fitting parameters are listed in Table 1. As for ternary mixtures, dry time data and their simulations for chips a, b, c are illustrated in Figures 5A, S3(A), S3(B), together with their fitting parameters summarized in supplementary Tables S2, S3 and S4, respectively, and color plots demonstrating process of the backward algorithm for all three chips are shown in Figure S3; the predicted concentrations on chip a are listed in Figure 5C and chips b, c in supplementary Table S5. In each drying experiment, after the IOF chip was infiltrated with different pure hydrocarbons or their mixtures by immersing the chip into liquids, the extra liquid on the top of the IOF was washed off by deionized water, and the IOF chip was immediately inserted into a chamber to prevent outside influence, such as air flows, on the evaporation rates. The chamber is made of 18 ACS Paragon Plus Environment

Page 19 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


two pieces of rectangular silicon and glass chip, of which three edges of the perimeters were bonded together by high temperature air set cement (Omega Bond 300 by OMEGA Engineering), leaving one small entrance for the IOF chip. The chamber was then placed on a heater with capacity to reach desired temperature within one second. The reflected light comprising information regarding the time-evolution of spectral intensity was collected and focused onto a fiber by a combination of several aligned lenses and further transferred into a spectrometer by the fiber. Between each experiment, the substrates are cleaned with IPA, acetone and water, and then dried in nitrogen to make sure no remaining liquids inside the inverse opal. ASSOCIATED CONTENT Supporting Information Photographs of drying process. SEM images of the IOF structures. Fitting parameters in modified Penman Equation. Practical and simulated drying times for different two-element mixtures. Figures of predicted concentration for three-component mixtures. AUTHOR INFORMATION Corresponding Author *[email protected]; [email protected] Author Contributions The manuscript was written through contributions of all authors. Funding Natural Sciences and Engineering Research Council of Canada (NSERC). U.S. Department of Transportation (Award number: DTPH5617C00002 and DTFR5315C00025). Notes The authors declare no competing financial interest 19 ACS Paragon Plus Environment


ACKNOWLEDGEMENTS The authors thank L. Meng for surface functionalization of IOF samples. REFERENCES (1)

Drews, A. W. Manual on Hydrocarbon Analysis. 1998, 1061 pages.

(2)

Blumer, M.; Guillard, R. R. L.; Chase, T. Hydrocarbons of Marine Phytoplankton. Mar. Biol. 1971, 8, 183–189.

(3)

Maynard, J. B.; Sanders, W. N. Determination of the Detailed Hydrocarbon Composition and Potential Atmospheric Reactivity of Full-Range Motor Gasolines. J. Air Pollut. Control Assoc. 1969, 19, 505–510.

(4)

Stockton, A. M.; Chiesl, T. N.; Scherer, J. R.; Mathies, R. A. Polycyclic Aromatic Hydrocarbon Analysis with the Mars Organic Analyzer Microchip Capillary Electrophoresis System. Anal. Chem. 2009, 81, 790–796.

(5)

Teng, S. T.; Williams, A. D.; Urdal, K. Detailed Hydrocarbon Analysis of Gasoline by GC‐MS (SI‐PIONA). J. High Resolut. Chromatogr. 1994, 17, 469–475.

(6)

Schoenmakers, P. J.; Oomen, J. L. M. M.; Blomberg, J.; Genuit, W.; van Velzen, G. Comparison of Comprehensive Two-Dimensional Gas Chromatography and Gas Chromatography - Mass Spectrometry for the Characterization of Complex Hydrocarbon Mixtures. J. Chromatogr. A 2000, 892, 29–46.

(7)

Van Geem, K. M.; Pyl, S. P.; Reyniers, M.-F.; Vercammen, J.; Beens, J.; Marin, G. B. OnLine Analysis of Complex Hydrocarbon Mixtures Using Comprehensive TwoDimensional Gas Chromatography. J. Chromatogr. A 2010, 1217, 6623–6633.

(8)

Mao, D.; Lookman, R.; Van De Weghe, H.; Van Look, D.; Vanermen, G.; De Brucker, N.; Diels, L. Detailed Analysis of Petroleum Hydrocarbon Attenuation in Biopiles by 20 ACS Paragon Plus Environment

Page 20 of 24

Page 21 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


High-Performance

Liquid

Chromatography

Followed

by

Comprehensive

Two-

Dimensional Gas Chromatography. J. Chromatogr. A 2009, 1216, 1524–1527. (9)

Han, M.; Assanis, D. N.; Jacobs, T. J.; Bohac, S. V. Method and Detailed Analysis of Individual Hydrocarbon Species From Diesel Combustion Modes and Diesel Oxidation Catalyst. J. Eng. Gas Turbines Power 2008, 130, 42803.

(10)

Vogel, A.; Cook, J. W. A Text-Book of Practical Organic Chemistry. Nature 1948, 161, 995–995.

(11)

Geddes, R. L. A General Index of Fractional Distillation Power for Hydrocarbon Mixtures. AIChE J. 1958, 4, 389–392.

(12)

Fan, X.; White, I. M. Optofluidic Microsystems for Chemical and Biological Analysis. Nat. Photonics 2011, 5, 591–597.

(13)

Huang, Y.; Pandraud, G.; Sarro, P. M. Reflectance-Based Two-Dimensional TiO_2 Photonic Crystal Liquid Sensors. Opt. Lett. 2012, 37, 3162.

(14)

Liu, Y.; Salemink, H. W. M. Photonic Crystal-Based All-Optical on-Chip Sensor. Opt. Express 2012, 20, 19912–19920.

(15)

Wu, D. K. C.; Kuhlmey, B. T.; Eggleton, B. J. Ultrasensitive Photonic Crystal Fiber Refractive Index Sensor. Opt. Lett. 2009, 34, 322.

(16)

Yang, J. C.; Ji, J.; Hogle, J. M.; Larson, D. N. Metallic Nanohole Arrays on Fluoropolymer Substrates as Small Label-Free Real-Time Bioprobes. Nano Lett. 2008, 8, 2718–2724.

(17)

Yanik, A. A.; Huang, M.; Artar, A.; Chang, T. Y.; Altug, H. Integrated NanoplasmonicNanofluidic Biosensors with Targeted Delivery of Analytes. Appl. Phys. Lett. 2010, 96, 21101. 21 ACS Paragon Plus Environment


(18)

Page 22 of 24

Escobedo, C.; Brolo, A. G.; Gordon, R.; Sinton, D. Flow-through vs Flow-over: Analysis of Transport and Binding in Nanohole Array Plasmonic Biosensors. Anal. Chem. 2010, 82, 10015–10020.

(19)

Nishijima, Y.; Ueno, K.; Juodkazis, S.; Mizeikis, V.; Misawa, H.; Tanimura, T.; Maeda, K. Inverse Silica Opal Photonic Crystals for Optical Sensing Applications. Opt. Express 2007, 15, 12979.

(20)

Zhao, X.; Xue, J.; Mu, Z.; Huang, Y.; Lu, M.; Gu, Z. Gold Nanoparticle Incorporated Inverse Opal Photonic Crystal Capillaries for Optofluidic Surface Enhanced Raman Spectroscopy. Biosens. Bioelectron. 2015, 72, 268–274.

(21)

Raymond, K. P.; Burgess, I. B.; Kinney, M. H.; Lončar, M.; Aizenberg, J. Combinatorial Wetting in Colour: An Optofluidic Nose. Lab Chip 2012, 12, 3666.

(22)

Burgess, I. B.; Koay, N.; Raymond, K. P.; Kolle, M.; Lončar, M.; Aizenberg, J. Wetting in Color: Colorimetric Differentiation of Organic Liquids with High Selectivity. ACS Nano 2012, 6, 1427–1437.

(23)

Zhang, W.; Gao, N.; Cui, J.; Wang, C.; Wang, S.; Zhang, G.; Dong, X.; Zhang, D.; Li, G. AIE-Doped Poly(ionic Liquid) Photonic Spheres: A Single Sphere-Based Customizable Sensing Platform for the Discrimination of Multi-Analytes. Chem. Sci. 2017, 8, 6281– 6289.

(24)

Aguirre, C. I.; Reguera, E.; Stein, A. Tunable Colors in Opals and Inverse Opal Photonic Crystals. Adv. Funct. Mater. 2010, 20, 2565–2578.

(25)

Fenzl, C.; Hirsch, T.; Wolfbeis, O. S. Photonic Crystals for Chemical Sensing and Biosensing. Angewandte Chemie - International Edition. 2014, pp 3318–3335.

(26)

Sedighi, A.; Qiu, S.; Wong, M. C. K.; Li, P. C. H. Dip-in Indicators for Visual 22 ACS Paragon Plus Environment

Page 23 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Differentiation of Fuel Mixtures Based on Wettability of Fluoroalkylchlorosilane-Coated Inverse Opal Films. ACS Appl. Mater. Interfaces 2015, 7, 28387–28392. (27)

Hong, J.; Park, D. H.; Baek, S.; Song, S.; Lee, C. W.; Kim, J. M. PolydiacetyleneEmbedded Microbeads for Colorimetric and Volumetric Sensing of Hydrocarbons. ACS Appl. Mater. Interfaces 2015, 7, 8339–8343.

(28)

Lee, J. H.; Fan, B.; Samdin, T. D.; Monteiro, D. A.; Desai, M. S.; Scheideler, O.; Jin, H. E.; Kim, S.; Lee, S. W. Phage-Based Structural Color Sensors and Their Pattern Recognition Sensing System. ACS Nano 2017, 11, 3632–3641.

(29)

Burgess, I. B.; Abedzadeh, N.; Kay, T. M.; Shneidman, A. V.; Cranshaw, D. J.; Lončar, M.; Aizenberg, J. Tuning and Freezing Disorder in Photonic Crystals Using Percolation Lithography. Sci. Rep. 2016, 6, 19542.

(30)

Penman, H. L. Natural Evporation from Open Water, Bare Soil and Grass. R. Soc. 1948, 193, 120–145.

(31)

Shuttleworth, W. J. Putting The “vap” into Evaporation. Hydrol. Earth Syst. Sci. 2007, 11, 210–244.

(32)

Thomson, G. W. The Antoine Equation for Vapor-Pressure Data. Chem. Rev. 1946, 38, 1– 39.

(33)

Guggenheim, E. A. The Theoretical Basis of Raoult’s Law. Transactions of the Faraday Society. 1937, p 151.

(34)

Li, S.-S.; Guan, Q.-Y.; Zheng, M.; Wang, Y.-Q.; Ye, D.; Kang, B.; Xu, J.-J.; Chen, H.-Y. Simultaneous Quantification of Multiple Endogenous Biothiols in Single Living Cells by Plasmonic Raman Probes. Chem. Sci. 2017, 8, 7582–7587.

(35)

Hatton, B.; Mishchenko, L.; Davis, S.; Sandhage, K. H.; Aizenberg, J. Assembly of Large23 ACS Paragon Plus Environment


Page 24 of 24

Area, Highly Ordered, Crack-Free Inverse Opal Films. Proc. Natl. Acad. Sci. 2010, 107, 10354–10359.


Inverse Opal Photonic Crystals as an Optofluidic Platform for Fast

Recommend Documents