Determination of Thermodynamic Properties of OLED Compounds
1. Determination of Thermodynamic
Properties of Organic Compounds
Team Serval: Logan Williamson, Owen Perlowski, Kristen Webster, Nick Kasper
CHE 255: Chemical Engineering Processes
December 13th, 2016
3. Current method:
Guess and Check = Time and Money
Problem
3
New
Material
Guess ideal
temp zones to
purify product
Train
sublimation
under vacuum
Successful
Separation?
(HPLC)
No Yes
4. Goals
Measure heat of vaporization
Run experiments under reduced pressure/temperature
Required accuracy: ballpark
Stay within budget/time constraints
4
6. Our Approach: Experiments
Thermogravimetric
Analysis
- Mass loss vs. temperature/time
- Clausius-Clapeyron equation
- No need to approach
degradation temperature
6
Differential Scanning
Calorimetry
- Differential heat flow vs.
temperature
- Lower pressure by sealing pans
under vacuum
Pressure reduction method
unsuccessful
Successful and promising
for application
7. Thermogravimetric Analysis (TGA): What is it?
http://www.tainstruments.com/tga-5500/ https://www.researchgate.net/figure/263958136_fig2_Figure-2-Schematic-of-the-TGA
7
8. TGA: Theory
8Source: Price, Duncan M. "Vapor Pressure Determination by Thermogravimetry."Thermochimica Acta 367-368 (2001): 253-62.
9. TGA: Experimental Outline
9
Determine Calibration
Constant (k)
Using k, Calculate Vapor
Pressures of Model
Compounds
Use Vapor Pressures to
Calculate Heat of
Vaporization of Model
Compounds
Compare Calculated
Values to Literature
Values
Use Method to Calculate
Vapor Pressures and
Heat of Vaporization of
Compound with No
Literature Data
11. TGA: Determining k for Each Compound
11
Anthracene
k = 887,652
Naphthalene
k = 1,503,009
12. TGA: Calculating Vapor Pressure Using
Optimized k
12
Using Least-Squares Regression:
Naphthalene
Temp (℃) Lit. p (Pa)
Avg.
Calculated p
(Pa)
Percent
Error
80 1000 828 17.2
93 1811 1672 7.7
98 2357 2036 13.6
103 2833 2418 14.7
115 4752 3754 21.0
Anthracene
Temp (℃) Lit. p (Pa)
Avg.
Calculated p
(Pa)
Percent
Error
227 7000 7898 12.8
237 9000 10,227 13.6
247 12000 12,051 0.4
257 16000 13,330 16.7
Optimal k = 1,183,875
Average % Error: 13%
13. TGA: Calculating Heats of Vaporization of Model Compounds
13
Naphthalene Anthracene
Calculated ΔHVap 49.3 kJ/mol
Literature ΔHVap 56.1 kJ/mol
Percent Error 12.1
Calculated ΔHVap 41.2 kJ/mol
Literature ΔHVap 59.2 kJ/mol
Percent Error 30.1
14. TGA: Heat of Vaporization is Independent of k
14
Naphthalene
kOpt = 1,183,875
ΔHVap = 49.3 kJ/mol
k = 1
ΔHVap = 49.3 kJ/mol
15. TGA: Calculating Heat of Vaporization of Unknown
Compound
15
Tri-p-tolylamine (TTA)
Temp (℃) Avg. Calculated p (Pa)
250 2692
280 5157
Calculated ΔHVap 52.2 kJ/mol
Molecular Weight: 287.4 g/mol
Using kOpt = 1,183,875
16. TGA: Summary
16
● Good for determining how long it will take to vaporize any compound at a
given temperature. This could help Molecular Glasses save time during
their purification of materials. TGA is versatile and can reach temperatures
> 1000℃.
● Vapor pressures can be estimated within ~20% error. One possible way to
reduce the error is to test more than 2 model compounds when regressing
for the calibration constant. Also test wider range of temperatures.
● Heats of Vaporization can be estimated using calculated vapor pressures.
Furthermore, these heats are independent of the calibration constant and
can therefore be calculated strictly from a TGA test and the Clausius-
Clapeyron equation.
19. Changes from Initial Build
Sealed areas that could
have allowed air leakage
with gorilla glue
Added cardboard blast
shield
19
20. Changes from Initial Build
Changed pressure gauge
after valve to a liquid filled
gauge to increase
readability
20
21. Vacuum Bagging Procedure
1. Cut bagging film into 3”x 8” sheets.
2. Place connector base and DSC pan with sample and lid on top on top of one
sheet.
3. Apply sealant tape around edges.
4. Place second sheet on top of the first and press down on tape to seal.
5. Connect connector to pump system and turn on pump.
6. Close valve 2 and allow pressure to build up within vessel.
7. Close valve 1 and open valve 2 to remove air from bag. 21
24. DSC
24www.pcbshop.prg
DSC Procedure was split into three
different sections, for the three different
pans we used:
1. Non-Hermetically Sealed Pans
2. Hermetically Sealed Pans under
vacuum
3. Hermetically Sealed pans not under
vacuum (1 atm)
The following slides will provide a simple
guide to each type of pan procedure
25. DSC - Procedure
(Non-Hermetic Pans Supplied by Mark Juba, Hermetically Sealed Pans provided
by Anthamatten Lab)
1. Place desired sample in pan. Weigh pan before and after.
a. If Hermetically sealed, use TA clamp to seal shut
2. Determine procedure in DSC program
a. Determine max temp/temp increase.
3. Input reference pan, along with pan with sample
4. Run sample
25
26. DSC: Results
Results for the DSC were in this form:
This graph shows us the Heat Flow (W/g) vs.
Temperature (C) .
As we know, the DSC records required heat for
phase change. This is expressed as a peak on
this graph.
By using manual integration present in the DSC
analysis software, we are able to find the precise
amount of heat required for this specific instance
of phase change. These values were compared
to literature for accuracy and error percentages
were analyzed calculated.
26
27. Calc H Fusion
(kJ/mol)
Calc H Vap
(kJ/mol)
Calc H Sub
(kJ/mol)
24.966 47.241 72.206
%Error 16.222 20.201 26.244
Calc H Fusion
(kJ/mol)
Calc H Vap
(kJ/mol)
Calc H Sub
(kJ/mol)
27.549 45.797 73.347
%Error 7.551 22.639 25.079
Anthracene Hermetic 1atm Trials
Anthracene Literature BP: 342C
Anthracene Literature MP: 218C
MP & BP Found at:
O'Neil, M.J. (ed.). The Merck Index - An Encyclopedia of Chemicals, Drugs, and Biologicals. Whitehouse
Station, NJ: Merck and Co., Inc., 2006., p. 111 27
Mass = 17.3mg
Mass = 9mg
28. Calc H Fusion
(kJ/mol)
Calc H Vap
(kJ/mol)
Calc H Sub
(kJ/mol)
27.549 46.778 74.327
%Error 7.551 20.983 24.078
Calc H Fusion
(kJ/mol)
Calc H Vap
(kJ/mol)
Calc H Sub
(kJ/mol)
25.251 47.793 73.044
%Error 15.265 19.268 25.389
Anthracene Hermetic Vacuum Trials
Anthracene Literature BP: 342C
Anthracene Literature MP: 218C
28
Mass = 14.8mg
Mass = 7.2mg
30. Calc H Fusion
(kJ/mol)
Calc H Vap
(kJ/mol)
Calc H Sub
(kJ/mol)
- - 98.188
%Error - - 0.294
Mass(mg) Calc H Fusion
(kJ/mol)
Calc H Vap
(kJ/mol)
Calc H Sub
(kJ/mol)
25.7 - - 93.537
%Error - - 4.456
Anthracene Non-Hermetic Trials
Anthracene Literature BP: 342C
Anthracene Literature MP: 218C
30
Mass = 13.4mg
Mass = 25.7mg
32. DSC: Summary
Saw no improvement with using vacuum bagging system to reduce pressure
inside the pan.
Increase in boiling point for sealed pans may tell us there is a small amount of
pressure buildup in our pans, which is counterproductive.
We observed literature consistency with non-hermetically sealed pans, although
it appears impossible to determine only heat of vaporization or only heat of
fusion for these trials.
No difference when using different mass amounts. Data is consistent regardless
of mass.
Observed relatively accurate heats for our compounds. This could be incredibly
32
33. Conclusion
Our goal was to determine the heat of vaporization of solid compounds under
reduced pressure for application to OLED materials
2 methods:
Differential scanning calorimetry (DSC) w/ vacuum sealing
Thermogravimetric analysis (TGA)
TGA and non-hermetically sealed DSC gave results most consistent with the
literature
No improvement with using vacuum bagging system to reduce pressure inside
the pan due to a buildup of internal pressure in the pans 33
34. Acknowledgments
We would like to thank everyone in the chemical engineering
department for allowing us this opportunity with a special
thanks to Professor Kelley, Mark Juba, Dave Weiss, Larry
Kuntz, Professor Tenhaeff, Cindy Fitzgerald, Rachel
Monfredo, our TA Robbie Harding, and Dawei Chen from the
Anthamatten Lab.
34
37. Additional TGA Analysis
37
Error Analysis:
Hypothesis: Noise in the data increases as the temperature increases.
Naphthalene Anthracene
Temperature [℃]
dm/dt[g/min]
Temperature [℃]
dm/dt[g/min]
38. References
Melting Point and Boiling Point of Anthracene:
O'Neil, M.J. (ed.). The Merck Index - An Encyclopedia of Chemicals, Drugs, and Biologicals. Whitehouse Station, NJ: Merck and Co., Inc., 2006., p. 111
Heat of Fusion & Vaporization, Anthracene:
Rojas, Aarón; Orozco, Eulogio, Measurement of the enthalpies of vaporization and sublimation of solids aromatic hydrocarbons by differential scanning calorimetry,
Thermochimica Acta, 2003, 405, 1, 93-107
Heat of Sublimation, Anthracene:
Oja, Vahur; Chen, Xu; Hajaligol, Mohammad R.; Chan, W. Geoffrey, Sublimation Thermodynamic Parameters for Cholesterol, Ergosterol, β-Sitosterol, and Stigmasterol, J.
Chem. Eng. Data, 2009, 54, 3, 730-734.
Thermogravimetric Methodology:
Price, Duncan M. "Vapor Pressure Determination by Thermogravimetry."Thermochimica Acta 367-368 (2001): 253-62.
38
Hinweis der Redaktion
We ball out
Molecular Glasses’ problem is that determining the properties of their new materials takes far too long and costs too much material.
They basically purify their sample under trial conditions and see how well it separates… then do it again… and again…
The goals of our project were:
Measure the heat of sublimation of model compounds
FInd ways to reduce pressure and/or avoid the high temps that degrade Molecular Glasses’ materials
The company isn’t looking for specific values, but a ballpark range for purification purposes
Just like all the other teams, we had a timeline of about 13 weeks and a budget of $500
(Nick) (I can talk about model compounds because it seems to lump in well with the previous slide)
Our two model compounds were naphthalene and anthracene. These compounds are high vapor pressure solids that work well to model the organic compounds that Molecular Glasses will need to test.
Anthracene is less volatile than naphthalene, both are less volatile than real OLEDs
We tested two experimental methods to determine heat of sublimation
One of the methods we used to determine vapor pressure and heat of vaporization of these model compounds is thermogravimetric analysis. This is a type of thermal analysis that measures mass loss as a function of temperature and time. It is a versatile instrument that can reach temperatures of over 1000 degrees celsius and is therefore commonly used to measure the thermal stability of chemical compounds. Our use of TGA focuses around a paper published by Duncan Price in which he uses a TGA under normal operating conditions to estimate the vapor pressure and heat of vaporization of various compounds.
Price’s methodology begins from the Langmuir equation for free evaporation. This equation has a couple assumptions, including that vaporization is a zeroth order process, meaning that rate of vaporization is independent of the amount of material, and that there is a uniform surface area for evaporation to occur.
Both the vapor pressure, p, and the vaporization coefficient, alpha, are unknown in this equation when the process is carried out under atmospheric conditions. Rearranging this equation and letting k = READ SLIDE and v = READ slide, we are able to obtain a linear relationship between vapor pressure and v with a slope of k, which is a calibration constant that is assumed to be independent of the chemical being vaporized. If we are able to determine this calibration constant, we will then be able to calculate vapor pressure using this value of k and v, which can be calculated from the results of a TGA trial.
Once vapor pressure is known, heat of vaporization can be determined using the Clausius-Clapeyron equation.
The experimental outline for the TGA method are as follows:
Various tests were conducted on naphthalene and anthracene and the using the results of the tests and the literature vapor pressures, a calibration constant for the machine was calculated.
This k value was used to calculate the vapor pressures of the model compounds
Once these vapor pressures were calculated, the Clausius-Clapeyron Equation was used to calculate the heat of vaporization of the model compounds.
These values were then compared to literature values to analyze the validity of the method.
The method was then repeated in order to calculate the vapor pressures and heat of vaporization of a compound with no known thermodynamic literature data.
Above is an example of the output of a single TGA trial, in this case naphthalene. As you can see the software very nicely allows us to calculate the rate of mass loss, which is denoted by the slope. It also confirms the assumption that rate of vaporization is independent of amount of material. These slope values, along with corresponding temperature and molecular weight of the given compound, are what was used to calculate v.
As was described on a previous slide, the first step in the analysis was to determine the calibration constant.
To do this we plotted literature vapor pressure values as a function of v for both naphthalene and anthracene. This resulted in a k value of 1,503,009 for naphthalene and 887,652 for anthracene. Clearly, these values are the same as was the assumption made when deriving the equations. We had to come up with a way to find an optimal k value to use for vapor pressure calculations.
After a lot of thought, it was decided to use a least squares regression of all the data (naphthalene and anthracene) , minimizing the error between the back-calculated vapor pressures and the literature values, to calculate an optimal k value. This optimized calibration constant was used to calculate the vapor pressures shown. As we can see, this results are reasonable. We are able to calculate average vapor pressures to within 21% of the literature values. The average error is 13%, which is lends some credibility to the method for determining vapor pressures. If more model compounds were tested this uncertainty could presumable be improved.
Using these calculated vapor pressures, the Clausius-Clapeyron equation was used to calculate the heat of vaporization of both naphthalene and anthracene. For those who may be unfamiliar with this formulation, a plot of the natural log of vapor pressure vs the reciprocal absolute temperature gives a linear relationship with a slope of heat of vaporization divided by the gas constant. Multiplying through by the gas constant gives us our heat of vaporization. As can be seen, the heat of vaporization of naphthalene was calculated to be 49.3 kJ/mol, resulting in a 12.1% error from the literature value. Likewise, anthracene had a calculated vapor pressure of 41.2 kJ/mol, resulting in a 21% error. As we see, our calculated enthalpies are less than the literature values, which could be the result of the Clausius-Clapeyron equation assuming that heats of vaporization are independent of temperature, which is an assumption that is made to make the calculations easier but is not exact.. Also, the level of variance in the anthracene data could be resulting in the greater percent error.
One of the most fascinating results of the project was that we were able to determine that the value calculated for heat of vaporization is in fact independent of the k value. As long as k is greater than zero, any value can be entered and the same heat of vaporization will be calculated, as can be seen here. When k is altered, the only thing that changes is the y-intercept. Therefore, if all you are looking to calculate is the heat of vaporization and don’t care about vapor pressures, you can simply run a TGA test on your compound and use the Clausius-Clapeyron Equation and boom you have your heat of vaporization. However, if you’re interested in determining vapor pressures, it is still necessary to find an optimal calibration constant.
Since the method seems viable, we went ahead and ran a few TGA tests on tri-p-tolylamine, an actual OLED compound, that has no literature vapor pressure and heat of vaporization data. At 250 degrees Celsius the vapor pressure was calculated to be 2692 Pa and at 280 degrees C it was 5157 Pa. The heat of vaporization of the compound was found to be 52.2 kJ/mol. Clearly, due to time constraints, very few temperatures were able to be tested. However, the idea behind the methodology seems to be viable.
Differential Scanning Calorimetry compares required heat needed for a pan with sample vs. a pan without sample to maintain constant increase in temperature.
The proposed idea of Lab Group Serval was to hermetically seal a DSC pan under low pressure, in an attempt to decrease observed boiling point of model compounds Naphthalene and Anthracene.
This would aid Molecular Glasses in their attempt to experimentally determine heats of sublimation/vaporization for their compounds in question
Anthracene 1atm Sealed
Anthracene 1atm Sealed
To visualize this effect, box and whisker plots were constructed with rate of mass loss being plotted as a function of temperature. As you can see, there does seem to be some correlation between temperature and degree of variability in the data, with the amount of spread increasing as temperature increases. However, we cannot confirm the temperature increase as the causation and further testing should be carried out on other heavy compounds that require high temperatures to see if this trend continues if this method if pursued. This is important to note because OLED materials tend to be involatile and therefore need to be purified at higher temperatures. This led to the problem of what value for k do we use to calculate vapor pressures and heat of vaporization?
