Theoretical Chemistry

Theoretical Chemistry

  • Muhammad Khattab added an answer:
    How to use the output of low level of theory calculation, as input for a higher level of theory, in details please?

    i read that i can use the checkpoint file of low level method to save time and getting a better accuracy.

    Muhammad Khattab · National Research Center, Egypt

    Thanks you all!

  • Shobeir K. S. Mazinani added an answer:
    How can I calculate the individual atomic polarzabilities in a molecule using molpro/gaussian?

    I have a molecule containing two atoms I want to calculate polarizability tensor for each atom in the molecule @ ccsd(T).

    Shobeir K. S. Mazinani · Arizona State University

    It's all about partitioning of the molecular polarizability. In my paper, Electronic transport through hydrogen bond we have used this technique. Please look at the introduction and the relevant references.

  • Tarik Ömer Oğurtani added an answer:
    How do you calculate ΔG of adsorbed species?

    Dear all,

    I am trying to calculate  ΔG of adsorbed species by using VASP. For this I need to calculate Zero point energies and Entropy of absorbed species. But I am confused with that. What type of calculations do I need to get ZPE and Entropy. Please help me in this case.

    Thanks in advance.

    anjibabu

    Tarik Ömer Oğurtani · Middle East Technical University

    Dear Anji, All these prototype shortwaves no matter how sophisticated they are,  one would have heal of a time to adapt them to the specific problems one faces in materials science. Because of that I have always developed my own tailor made codes in first MatCad environment and then switch to  FORTRAN,    including the mathematical subroutines even though some of them are available in IMSL etc. 

    One easily organize 

  • Iresh Gallindawatte asked a question:
    How do I determine KPoint values for a molecule for DFTB+ calculations?

    I need to know how to determine KPointAndWeight to fill out the input .hsd file.Anybody know about this? your consideration is highly appreciated.

  • Umadevi palanivel added an answer:
    How can I obtain B in calculations of frequency-dependent hyperpolarizability from its tensors?

    I am using cam-b3lyp/6-31g(d) POLAR cphf=rdfreq in gaussian 09 to
    calculate the frequency-dependent hyperpolarizability. Finally I got 18 beta tensor elements.
    Do you know exactly what these 18 tensors are?
    and how i can obtain B(-2w,w,w) from these 18 tensor?
    thank you very much

    Umadevi palanivel · Bharathiar University

    thanks for your prompt reply. squaring of the <beta zzz> doesnot gives the HRS value. i tried to reprodeuce some work in IJQC. but i cant reproduce the data by just squaring and adding zzz, xzz. 

  • Eugene F Kislyakov added an answer:
    Software to lower the symmetry of molecules?

    I am having difficulty in lowering the symmetry of a big molecule. My molecule has D3h symmetry, but I want to calculate it in C2v symmetry. Does anyone know the proper software with which to do that?

    Eugene F Kislyakov · Belarusian State University

    The same group can mean different geometries. It is more convenient to define geometry, than group. Standard procedure.

  • What does erroneous write stand for?

    Working with silicon two-layered clusters, I get an error message that goes like:

    Calling FoFCou, ICntrl= 3107 FMM=T I1Cent= 0 AccDes= 0.00D+00.
    Calling FoFCou, ICntrl= 3107 FMM=T I1Cent= 0 AccDes= 0.00D+00.
    End of G2Drv F.D. properties file 721 does not exist.
    End of G2Drv F.D. properties file 722 does not exist.
    End of G2Drv F.D. properties file 788 does not exist.
    IDoAtm=11111111111111111111111111111111111111111111111111
    IDoAtm=11111111111111111111111111111111111111111111111111
    IDoAtm=11111111111111111111111111111111111111111111
    Differentiating once with respect to electric field.
    with respect to dipole field.
    Differentiating once with respect to nuclear coordinates.
    CalDSu exits because no D1Ps are significant.
    There are 435 degrees of freedom in the 1st order CPHF. IDoFFX=6 NUNeed= 3.
    435 vectors produced by pass 0 Test12= 1.13D-12 1.00D-09 XBig12= 1.46D+04 1.75D+01.
    AX will form 72 AO Fock derivatives at one time.
    435 vectors produced by pass 1 Test12= 1.13D-12 1.00D-09 XBig12= 6.83D+03 2.12D+00.
    Erroneous write. Write -1 instead of 509295232.
    fd = 4
    orig len = 650011392 left = 509295232
    g_write

    I am using pbe0/6-31G** requesting a freq calc and adding the xqc on the scf, annexed are my files. It's not the first time I get this message for I've adjusted the system several times with no result

    Thanks!

    Best regards

    Juan-Sebastián Gómez-Jeria · University of Chile

    Use the option int=nofofcou and tell me it it worked or not. Regards, Juan

  • Bojidarka B. Ivanova added an answer:
    What is the difference between AM1 and PM3 methods?

    What is the difference between AM1 and PM3 methods in computational chemistry and when is it better to use one over the other?

    Bojidarka B. Ivanova · Technische Universität Dortmund

    Mr./Miss/Mrs. Albar,

    The formalism is the same, but the parameters are different.
    Please find in this topic the following refs:

    [1] M. Dewar, W. Thiel, Ground States of Molecules. 38. The MNDO Method.
    Approximations and Parameters, J. Am. Chem. Soc. 99 (1977) 4899 - 4907
    [2]M. Dewar, E. Zoebisch, E. Healy, J. Stewart, AM1: A New General Purpose Quantum Mechanical
    Molecular Model, J. Am. Chem. Soc. 107 (1985) 3902 - 3909

    [3] J. Stewart, Optimization of Parameters for Semiempirical Methods
    I. Method, J. Comput. Chem. 10 (1989) 209-220

    In addition you can find their application in vibrational spectroscopy and the correlations obtained for few examples (attachment).

  • Mumtaz Danish added an answer:
    Are there any examples of the cyclic reversible reaction network?

    Are there any examples of the following cyclic reversible reaction network?

    A <---> B; B <---> C; C <---> A

    Does any practical reaction follow this network or is this merely a theoretical possibility?  

    Mumtaz Danish · University of Karachi

    Dear Stegmaier thank you very much for this valuable suggestion about conformers.

  • Kaking Yan added an answer:
    How can I exchange a PF6- anion for Cl- in a ruthenium(II) complex?

    Is there any simple method to exchange PF6- counterion for some smaller group (as Cl-)? I need to increase water solubility of ruthenium(II) polypyridyl complex.

    Kaking Yan · University of California, Berkeley

    Dear Michał,

    I have recently come across a paper from J. M. Klein Gebbink on Chem. Eur. J. on exchanging PF6- counterion to Cl- on bimetallic Ru-Pd pincer compounds using Amberlite ion-exchange resin with NaCl as the Cl source. This might be a good starting point to check out. 

    chem.200900117

    Chem. Eur. J. 2009, 15, 3340 – 3343.

  • A. M. Verma added an answer:
    Does anyone here know how to generate Frank-Condon spectrum in GaussView?

    I use the 3 step approach and the error happens in the third one.The frequencies and checkpoints for the initial and final states are well generated but an error happens in the third step. 

    Sometimes generating the Frank-Condon spectrum starts and after a few lines the error happens and in most of the cases it does dot even start.

    I wonder if it happens because of a problem in entering the commands or there's a problem with checkpoints. A command sample that has properly worked is appreciated.

    Thank you in advance for your help and suggestions.

    Sara

    A. M. Verma · Indian Institute of Technology Guwahati

    Dear Sara Farahani

    did you go through with this link?

    http://faculty.ycp.edu/~jforesma/workshop/franck_condon/franck_condon.html

    this might help you...

    If you have gone through already then I would need the .com and .log files

    thanks

  • Davood Nori-Shargh added an answer:
    Does anyone know of software that can automatically determine the term symbol of a molecule?

    Is there any way to predict a term symbol for bigger molecules in both ground and excited states using software?

    Davood Nori-Shargh · Islamic Azad University

    The symmetries of the molecular orbitals can be obtained by means of GAMESS-US, GAUSSIAN and other programs but it depends on the symmetry of the structure that you are considering for the corresponding molecule. Therefore, the determinant step in the determination of the molecular orbital symmetries is the point group of the molecular structures. For the different geometries of a molecule, you will find the different molecular orbital symmetries.    

  • Ritika Joshi added an answer:
    Why does the lifetime of a probe decrease with increase in polarity and proticity of the solvent especially upto MeOH??

    This is the situation with one of the molecules I am studying. The decay curve is biexponential and fits the best for presence of two lifetimes.The molecule is showing geometrical relaxation to planarity as well as ICT in the excited state.

    While the lifetime is showing a sudden increase in water after MeOH.

    Thanks in advance

    Ritika Joshi · Visvesvaraya National Institute of Technology

    Hello Dr. Peter,

    Yes, you are very correct, the energy levels of various ES of my molecule changes with polarity of solvent.

    The LE or the non planar ICT state is showing an increase in lifetime from hexane to water (the percentage contribution being around 98-99% from hexane to MeOH while it being 15% in water)

    While, the planar ICT state is showing a lifetime decrease with increasing polarity n proticity of solvents   ie upto MeOH( percentage contribution being 1 or <1%). But the same species turns out to have 86% contribution in water with a sudden increase in lifetime.

  • Paul Jerabek added an answer:
    Can I get the same transition state for the different products of the same reactants?

    I have a SN2 reaction which may give two different products from the same reactants. Is it possible to get the same transition state structure for both of these reactions?

    Please let me know if I am thinking in the wrong way..

    Thanks and Regards

    Paul Jerabek · Philipps University of Marburg

    Hello Anjali,

    you have to perform an IRC calculation in order to "prove" that your found TS connects the reactants with your product. After you followed the IRC path forwards and backwards for several points (10-20 should be sufficient) you optimize the geometries and observe where the structures are going. In the best case you get your reactants and your product and you're done. If not, you have to find another TS and try again.

    In my understanding it should not be possible to have the exactly same TS for two different products. However, it could be that two TS are very similar in appearance and energy.

  • Bartosz Trzaskowski added an answer:
    How can one study confinement using a defined external potential in gaussain09 or any other software?

    J. Chem. Theory Comput.2013, 9, 3463−3472 like in this they studied the confinement using external potential

    Bartosz Trzaskowski · University of Warsaw

    You could also try something different (depending on what you really want to study). We did once a study of reaction pathway in gas phase and then compared it to the same reaction taking place inside a nanotube. This was using a very simple model: we used 2-layer ONIOM in Gaussian and the entire nanotube was modelled at the MM level: so the only effect of the nanotube was the confinement, but no electronic effects originating from the nanotube-reactants were taken into account.

    It's probably too simple for many applications, but the implementation in Gaussian (or any software with ONIOM or QM/MM capabilities) is straightforward.

  • Atul Sharma added an answer:
    What solution will I get adding 75:15 mm of EDC and NHS respectively in water?

    N-hydroxysuccinimide(NHS) and EDC(1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide) are expected to be water soluble.But my solution gave semi-ester form with  two immiscible portion of solid and water.If anyone is working with this system, please describe me the final form solution.

    Atul Sharma · BITS Pilani, K K Birla Goa

    There is no role of temperature. You can thiolate your metal surface and afterward use the EDC:NHS. But make sure do not keep it for longer time. One precaution you can use, keep it in desiccator after adding EDC: NHS mixture.

  • Maria Javaid added an answer:
    Can anyone help with an error message (convergence failure) in Gaussian 09?
    I'm working in Zr cluster system using DFT B3LYP calculation in Gaussian 09. I get an error message convergence failure. I tried simple basis set sto-3g basis set calculation checkpoint file moving to larger basis set DFT B3LYP but also get the same error. Any suggestions?
    Maria Javaid · RMIT University

    Hi Noor....

    I am facing problem with a gaussian09 convergence failure. I am trying to run monolayer MoS2 to see its energy gap with BHandHLYP/LANL2DZ. I got convergence up to 48 atoms but I want to get this for bigger systems like for 75 atoms, energy is not converging and I want to go to several hundred of atoms.

    I tried couple of things like scf=qc, fermi, conver=5, maxcycle=512. I tried to change initial guess= core, huckel. I distorted geometry a bit. I also tried scf=NoIncFock. None of them is working. Can you suggest anything please?

    Thanks

  • Martin Schwarzer added an answer:
    Why does my IRC Calculation in G09 only produces 1 point in the reverse direction but produces 9 points in the forward direction?

    My input file is as follows:

    %chk=qst3c.chk
    # B3LYP/cc-pVTZ irc=(CalcFC, maxpoints=10,stepsize=8) geom=allcheck

    Transition One IRC

    0 1

    The output file produces 9 points in the forward direction and only 1 in the reverse direction instead of 9. Any ideas why?

    Martin Schwarzer · Ibaraki University

    Without more information it is not easy to pinpoint to the real problem here. Depending on surrounding area on the potential energy hyper-surface this might just be what is supposed to happen. Here are some clues, what you can try.

    First of all reoptimize your transition state, using a finer grid and tighter cutoffs:

        %oldcheck=qst3.chk

        %chk=opt.chk

        #B3LYP/cc-pVTZ geom=allcheck guess=read opt(TS,calcfc,tight) int(ultrafinegrid)

    Then you should verify, that your transition state is actually the one you are looking for. Run a pure frequency calculation:

        %oldcheck=opt.chk

        %chk=freq.chk

        #B3LYP/cc-pVTZ FREQ geom=allcheck guess=read int(ultrafinegrid)

    Uncouple your IRC calculations, run one in forward and one in reverse direction, read in the force constants from the previous calculation. Use more points, more cycles, and set a different stepsize (see below):

        %oldcheck=freq.chk

        %chk=rev.chk

        #B3LYP/cc-pVTZ IRC(RCFC, Maxpoints=30, MaxCycle=300, StepSize=XXX, forward) geom=allcheck guess=read int(ultrafinegrid)

    You might want to try a stepsize quite small, like 3-5 for a very steep PES. If you have a very flat PES you have to increase this value, try 20 or 30. You might have to do this a couple of times until you find enough points to continue. (Usually it's a good guess to start with a small n.)

    You can set the convergence criteria with IOP(1/7=n). n=200-300 is reasonable. Smaller numbers mean tighter criteria. Tight criteria might be necessary for a shallow PES.

    If it still fails, try the old optimiser IRC(GS2). 

    Doing IRC calculations usually needs a lot of hand on work and experience. This is not straight out of the box. Practise it with small molecules to get a feeling.

    ***

    The procedure %oldchk copies the checkpoint file to the new location specified in %chk, You do not have to do this manually, thus avoiding possible errors.

    ***

    A note on your input. When you invoke geom=allcheck, you should not specify title and the charge/spin sections. They will be ignored anyway. If you invoke geom=check then you must specify these lines. This is a safer bet, since you exactly specify what is being done.

    ***

    I have included a link, that helped me, as it gathered the most important keywords for these kind of calculations.

  • Britney Stewart added an answer:
    Why does my IRC calculation in G09 only outputs 1 step in the forward and reverse direction?

    My input file is as follows:

    %chk=qst3c.chk
    # B3LYP/cc-pVTZ irc=(CalcFC, maxpoints=100,stepsize=10) geom=allcheck

    Transition One IRC

    0 1

    From my understanding, the default maxpoints is 10 and so is the step size. So I initially set the maxpoints at 10 with a stepsize of 1, however, it still only produced one step in both directions. Changing the maxpoints to 100 and the stepsize to 10, there was the same result. Is there something I may be doing wrong?

    Thanks in Advance

    Britney Stewart · Coastal Carolina University

    I changed the maxpoints=10 stepsize=8. It now produces points in the forward direction but only outputs 1 point in the reverse. Any ideas why?

  • Martin Schwarzer added an answer:
    How can I fix the convergence failure error in gaussian calculation in using PBEPBE set?

    Hi all,

    I tried to run a calculation on Gaussian with the PBEPBE basic set as follow:

    +++

    # PBEPBE/6-31g opt=(calcfc,noeigen) freq scf=(qc,tight,maxcyc=200)

    +++

    It's okay I can optimize the radical with no problem. However, if I continue to run as follow:

    # ROPBEPBE/6-311++G(d,p) guess=read geom=allcheck scf=(tight,maxcyc=20
    0)

    +++

    I have a convergence failure problem as follow:

    +++

    Annihilation of the first spin contaminant:
    S**2 before annihilation 0.7500, after 0.7500
    Convergence failure -- run terminated.
    Error termination via Lnk1e in C:\G09W\l502.exe at Sat Jan 31 07:45:38 2015.

    +++

    Please tell me how to fix this? Thanks in advanced.

    Martin Schwarzer · Ibaraki University

    The mentioned BMK, M06, PBEPBE, HSEH1PBE, and TPSSTPSS are density functionals and not basis sets. The basis sets you are using are Pople's basis sets 6-31G, etc. 

    Not all density functionals perform the same and there are different levels of accuracy. PBEPBE (or short PBE) is a pure functional, while M06 is a meta-hybrid functional. I can recommend the book "A Chemist's Guide to Density Functional Theory"
    by Wolfram Koch, Max C Holthausen to get a deeper insight to what that actually means. 

    Sometimes calculations fail because the electronic structure is very complicated, sometimes they succeed, but are meaningless. It is important to check orbitals and occupancy and see if different methods give different results.

    I do not know what happens in the code of Gaussian, so I really cannot tell you what went wrong at what point. I also do not know enough about your system to make any predictions.

  • Ramesh Chitumalla added an answer:
    Can someone help me in finding transition state?

    I'm performing a calculation of transition states for a free radical chain reaction (self accelerating reaction) . The chain reaction consists of 8 steps. I have calculated transition states for 6 steps using QST3 method in G09. But I’m not able to find the transition states for 1st and 8th steps. These two steps are of high activation energy, according to the literature. The first step (initialization reaction) is about breaking a peroxide bond of an aromatic hydro peroxide, thus forming a hydroxyl radical and aryloxy radical. Second step is release of methyl radical from the aryloxy radical so as to produce a stable closed shell product. Second step is a exothermic reaction with very less activation energy (6 kj/mol). The aromatic hydro peroxide has six rotamers, and this is creating problem in finding the transition state. whatever the initial guess I’m providing, most of the time output is a transition state regarding the conversion of one rotamer to another. I have also tried relaxed potential energy surface scan of peroxy bond and took initial guess from this scan for qst3. Then the resulting transition state is about release of a methyl radical along with hydroxyl radical. I tried freezing the bonds, but could not get the required transition state.
    So my questions are:
    1) Is the transition state for first step is very week, and that’s why im not able to find it?
    2) Is low activation energy, and high exothermicity of second step reason for why I’m not able to find the transition state for first step?
    Any help would be greatly appreciated.
    Thanks in advance.

    Ramesh Chitumalla · Pusan National University

    Thank you Dear Prof. Peter Daminan Jarowski 

  • Sascha Nehr added an answer:
    Is it possible to break chemical bonds in a cl2 molecule using photolysis?

    if we search about bond energy of cl2 reach to 243 kJ/mol (http://en.wikipedia.org/wiki/Halogen) and by convert unit and mathematical calculation and convert to nm of light (http://www.webqc.org/unitconverters-js.php) find the 493 nm and this wavelength is Related "bluish green" (493–498 nm) light. 
    if this light radiate to the cl2 gas what react is predictable for this conditions .

    Sascha Nehr · Verein Deutscher Ingenieure e. V.

    Full spectral information on gaseous molecules are available under:

    http://satellite.mpic.de/spectral_atlas

  • Lucas Fagundes Esteves added an answer:
    Has anybody had problems with IRC calculations within a PCM model?

    Dear Coleagues,

    I am investigating the hydrolysis process of phosphate esters using PCM model to simulate the surroundings of the reactive system. The process is completely described in gas phase, however in aqueous phase I have had some problems with the Intrinsic Coordinate Reaction (IRC) analysis. The behaviour of the curve was not normal and two main problems were observed:
    1 – When I try to obtain the IRC in two outputs, by specifying the forward and reverse keywords in the calculation line, what I get is the same path, i.e., the reactants reach the transition state (in the forward calculation) and after that the transition state reach the reactants once again (in the reverse calculation).
    2 – After this problem I tried to change the IRC parameters and I put a different stepsize and a diferent maxpoints. In this situation the change of step size were not regular and in one extremity of the curve the IRC points accumulate in the same region, where the stepsize remained practically unchanged.
    The set of keywords were:
    # irc=(calcfc,maxpoints=60,forward,stepsize=20,maxcyc=200) scf=(maxcycle=200) b3lyp/gen gfinput pseudo=read scrf=(pcm,read)
    At the end of input:
    Eps=78.39
    Radii=uahf
    Has anybody here had a similar problem in IRC calculations using the continuum approach?
    I really appreciate any help you can provide.

    Lucas Fagundes Esteves · Federal University of Juiz de Fora

    Dear Mihai, thank you for your suggestions. I spend the weekend analizing my structures and your intuition was right, the IRC for normal mechanism is wrong. The TS is right but the IRC was conducted assuming the gas phase as the surroundings, and the TS strucutre was obtained within the continuum media. So I remake the calculations using now the IRC in aqueous phase:

    # irc=(calcfc,maxpoints=60,forward,stepsize=20,maxcyc=200) scf=(maxcycle=200) b3lyp/gen gfinput pseudo=read scrf=(pcm,read)

    The problems with points accumulating in the extremity of the curve has appeared again, but the calculation is still running. I will wait until it finishes to comment about the results. 

    Once again thanks for the helpful assistance

    Dear Prerna, unfortunately i did not find any solution until now, but I am working hard to understanding what is happening with IRC calculations using PCM. I saw a comment in Joaquim Barroso's Blog about an incompatibility between IRC and PCM in Gaussian 03 (though I'm using G09), but he says that it is not true for Gaussian 09, where there is a new algorithm for PCM calculations.

    Blog link:
    http://joaquinbarroso.com/2009/09/07/pcmg03/#comment-251

    I am looking for similar problems in CCL, but I did not find anything similar until now. 

  • Can anyone tell me what is the best basis set to run polycarbonate samples?

    I am working on the polycarbonate compound. I would like to know which basis set will give me good data closer experimental values of frequency, geometrical parameters, reactivity details and other related values.

    Juan José Gamboa-Carballo · Instituto Superior de Tecnologías y Ciencias Aplicadas

    You should try with a TZV Pople's basis set, some times they are more accurate and lighter than Dunning's. I sugest not to use MP2 if you want termodinamics, rather use CCSD(T)

  • Robert J Paruch added an answer:
    How do you output the trajectory file in MD simulation?

    Sorry about the confusing title. 

    I am talking about the output frequency to write the trajectory file. Normally, the time-step in MD simulation is 1 or 2fs. And I am using 1ps/frame to output the trajectory file. The file size is big. And when I analysis these files. I choose 10ps/frame. I found it's hard to find difference to analyze the file between 1ps/frame and 10ps/frame. So should I change the frequency to write the trajectory file to 10ps/frame to save space? 

    How often do you guys store the trajectory files?

    Robert J Paruch · Pennsylvania State University

    One can also think of a variable output frequency if this is appropriate for the system and kind of simulation. For example, output frequency has to be high during the high energy stage of the simulated proces, when particles move quickly. Then, as the system equilibrates, the output frequency can be reduced. It helps to decrees the size of the output file, keeping the relevant information.

  • Ron Shepard added an answer:
    How do we calculate energies with multireference configuration interaction method in Gaussian 09?

    I try to calculate the energy with multireference configuration interaction method of certain system using gaussian 09 packet, but I do not know what is the keyword.

    Ron Shepard · Argonne National Laboratory

    COLUMBUS does MRCI, and it is also free. https://www.univie.ac.at/columbus/

  • Eric B. J. Harris added an answer:
    Why carboxyl group is inactive in Diels-Alder reaction?

    Are there any examples of Diels-Alder reaction in which carboxyl group (in acids or esters) acts as a part of a dien or a dienophile? What are the reasons of its inactivity in this reaction?

    Eric B. J. Harris · Australian National University

    Yes, there are examples of methylene barbituric acid derivatives and methylene meldrum's acid derivatives undergoing hetero Diels Alder reactions as a diene component. These systems are strongly electronically activated and sterically fixed in cisoid conformation.

    Cinnamates and similar systems have also been used but often require an electron rich dienophile and sometimes occur via step-wise rather than concerted process.

    The major limitation for the use of esters and carboxyl groups is that they are not electrophilic enough to make good dienophiles and possess poor orbital overlap on the carbon atom in addition to suffering from higher steric requirements.

  • Partha Sarathi Sengupta asked a question:
    By which methods can we get data on Fe(II) -C bonds in [Fe(CN)6](4-), and whether it is longer than that in [Fe(CN)6](3-).

    Fe(II) -C bonds in [Fe(CN)6](4-) is longer than that in [Fe(CN)6](3-). By which methods we get these data?

About Theoretical Chemistry

The use of mathematical and computational methods to build theories explaining observations of chemical reactions, such as quantum chemistry.

Topic Followers (14768) See all