Estimation of Hydrogen Bond Distribution in Coal

8/18/2019 Estimation of Hydrogen Bond Distribution in Coal

1/12

Estimation of Hydrogen Bond Distribution in Coalthrough the Analysis of OH Stretching Bands in Diffuse

Reflectance Infrared Spectrum Measured by in-SituTechnique

Kouichi Miura ,* Ka zuhiro Mae, Wen Li, † T ak um i K us ak aw a,

Fumia ki Morozumi, a nd Akiko Kuma noDepartment of Chemical Engineering, Kyoto University, Kyoto 606-8501, Japan

Received August 7, 2000. Revised M anuscript Received M arch 8, 2001

A new method wa s presented t o est ima te the st rength distribution of hydrogen bonds in coal.The h ydrogen bonds include the coal intra molecule hydrogen bonds and coal-water hydrogenbonds formed by hydroxyls in coal. The method analyzes the FTIR spectrum ranging from 2400to 3700 cm-1 obtain ed using the in-situ diffuse reflectan ce IR Fourier tra nsform (DRI FT) techniquewith neat , undiluted, coal sa mples. The FTIR spectra during t he hea t-up of eight coals (sevenArgonne premium coals and an Australian brown coal), an ion-exchange resin, and a lignin weremeasured every 20 °C from room temperature to 300 °C. Each spectrum was divided into six

hydrogen-bonded absorption bands by a curve-resolving method, then the amount of hydroxylscontributing t o each hydrogen bond w as est imated by B eer’s law by using different absorptivityfor each ba nd. The strength of each hydrogen bond w as est imated using a relat ion presented byD ra g o et a l . t h a t is k n own a s on e of t h e “ l in ea r e n t h a lpy-spectroscopic shift relations”. Usingthis a na lysis method, changes in hydrogen bond distributions (HB D) with increasing temperatur ewere successfully est imated for all th e samples exam ined. By ut ilizing the H BD the changes ine n t h a lp ie s a s s o c ia t e d wit h t h e de s o rp t io n o f a ds o rbe d wa t e r , t h e g la s s t ra n s i t io n , a n d t h edecomposit ion of COOH groups were well est imated. Only FTIR spectra measurements werefound to be enough to obtain such enthalpies. This greatly simplified the calculation procedurea n d in cre a s ed t h e a c cu ra c y of t h e e n t h a lpies . Th e v a l idit y of t h e p rop os ed in s i t u F TI Rmeasurement method and the analysis method for obtaining HBD was well clarif ied.

Introduction

I t is believed tha t f ive associat ive intera ct ions existin coals. 1 These a re covalent bonds, ionic linkages,hydrogen bonds, van der Waals forces, and π -π interac-t ion s . O f t h e f iv e a s s ocia t iv e in t era c t ion s h y drog enbonds, one of noncovalent associative interactions, arebelieved to play a key role to keep the m a cromolecula rstructure of low ra nk coals,1-7 an d hence utilizat ion oft h e l ow r a n k coa l s i s s ig n if ica n t l y a f f ect e d b y t h eh y dro g e n bo n din g . Sin c e a la rg e a mo u n t o f wa t e r isre t a in e d in t h e lo we r ra n k c o a l du e t o t h e h y dro g e nbonds, drying is one of draw backs for the ut ilizat ion ofthe coal. Hydrogen-bonded carboxylic a nd hydroxylic

g r ou ps i n t h e coa l cr os s -l in k a t t h e e a r ly s t a g e of

pyrolysis, which affects significantly the subsequentma in pyrolysis rea ct ions.8 Usua lly the cross-linkingreactions suppress the forma tion of volat iles. In solventextra ct ion of coal and/or t he liquefact ion of coal theinteract ions of coal surface and solvents play a crucialrole. From a fundamental viewpoint, physical propertiess u c h a s t h e g la s s t ra n s i t io n t e mp e ra t u re a re s ig n if i-ca n t ly a f f ect e d by t h e e xis t en ce of h y drog e n bon ds .Therefore it is very important to quantify the hydrogenbon ds in coa l f rom bot h p ra ct ica l a n d f u n da me n t a lviewpoints. Quantificat ion mea ns t he est imation of theamount and the strength of each hydrogen bond.

S e ve ra l a t t e m pt s h a v e b e en m a d e t o e s t im a t e t h ehydr ogen bonds in coal using va rious methods, including

solvent swelling phenomenon of coal, DSC, NMR, andFTIR. Experiments and analyses of solvent swelling ofc o a l h a v e be e n p e rf o rme d in re la t io n t o t h e ro le o fnoncovalent bonds in coal. One of the pioneer works ist h e o n e pe rforme d by Sa n a da a n d H o n da .9 They havee st ima t e d t h e de ns i t y of cros s -l in k s a n d t h e me a nmolecular weight between two adjacent cross-links byapplying the Flory-Rehler theory for polymers to the

* Corresponding author. Tel. +

81-75-753-5578. F a x. +

81-75-753-5909. E-mail: miura @cheme.kyoto-u.a c.jp.† On leave from Institute of Coal Chemistry, Chinese Academy of

Science, P.O. Box 165, Taiyuan, Shanxi 030001, P. R. China.(1) Larsen, J . W.; G reen, T. K.; K ovac, J . J . O r g . C h em . 1985, 50 ,

4729-4735.(2) Larsen, J . W.; Sha wver, Energy Fuels 1990, 4 , 74-77.(3) Lucht, L. M.; Peppas, N. A. Fuel 1987, 66 , 815-826.(4) Lucht, L. M.; Peppas, N. A. J. Appl. Polym. Sci. 1987, 3 3 , 2777-

2785.(5) Suuberg, E. M.; Otake, Y.; Sezen, Y. D i f f u s i o n i n C o a l s ; D O E

Rep ort (DOE /P C/80527-12), 1990.(6) Aida, T.; Squires, T. G. Prepr. Pap.s Am. Chem. Soc., Di v. Fu el

Chem. 1985, 30 , 95-102.(7 ) Mi ura, K. ; Ma e , K. ; Take be , S. ; Waki y asu, H.; Hashi moto, K.

Energy Fuels 1994, 8 , 874-880.

(8) Mae, K.; Maki, T.; Okutsu, H.; Miura, K. Fuel 2000, 79 , 417-425.

(9 ) Sa nada , Y. ; Honda, H. Fuel 1966, 45 , 295-300.

599E n er g y & Fu el s 2001, 15 , 5 99-610

10.1021/ef0001787 CC C: $20.00 © 2001 American C hemical SocietyP ubl ish ed on Web 04/13/2001


2/12

swelling of coal by pyridine vapor. Since this pioneerwork, many works have been presented and they havecontributed to clarify the coal structure. Recently Larsenet a l. , for exam ple, ha ve presented a method to estimat ethe number of hydrogen-bond cross-links in coal byutilizing the swelling of coal with polar solvent dilutedby nonpolar solvent.10 This method clarified that therea re a bout 1.7 hydr ogen-bond cross-links per 100 carbonsin Illinois No. 6 coal and 0.3 hydrogen-bond cross-links

per 100 carbons in Pittsburgh No. 8 coal. However, noinforma tion has been given on the strength dist ributionof hydrogen bonds in coal from solvent swelling exami-nations.

Direct or indirect est imation of heat accompanyingthe formation or disruption of hydrogen bonds is anotherme t h o d t o e s t ima t e t h e s t re n g t h o f h y dro g e n bo n ds .Lucht et al . ,11 Mackinnon et al . ,12 a n d Y u n et a l .,13 forexample, examined the glass transit ion of coal by theD SC me a s u reme n t . M a c k in n on e t a l . obs erv ed t h esecond-order phase glass transit ion at around 110 °Cduring the heating of various coals. However, directcalorimetric measurements have been in general notsuccessful, because m any factors have t o be taken int o

account to convert the experimentally obtained heat intothe str ength of hydrogen bonds.

H e a t s o f we t t in g o f c o a l in s o lv e n t s s u c h a s wa t e r ,methanol, etc. , have been measured calorimetrically toelucidate oxygen functional groups in coal. As early asin 1957 Iyenger and Lahiri tried to correlate the heatof wet ting a nd th e energy of hydrogen bonding betw eencoa l ox y g en f u n ct ion a l g rou p s a n d me t h a n o l.14 Tocalculate the energy of hydrogen bonds, they a ssumedtha t each COOH group is capable of hydrogen bondingw i t h t w o C H 3O H m o l e c u l e s a n d t h a t O H a n d CdOg rou p s a re e a c h ca p a ble o f h y dro g en bon din g w it ho n e CH 3OH molecule, respectively, and assigned thefollowing hydrogen bond energy value for each hydrogenbond formed by these functional groups: COOH: 8.2kca l/mol , CdO: 7.0 kcal/mol, and OH: 6.2 kcal mol.15

B y f u r t h e r a s s u m i n g t h a t a l l CdO groups are eitherchelated or interlocked w ith t heir corresponding numberof OH groups and therefore not available for hydrogenbonding wit h wa ter, which wa s deduced from the strongIR absorption band near 1600 cm-1 in all coals, theyfound a remarkable stat ist ical agreement between theheat of wetting a nd the t otal energy of hydrogen bondingcalcula ted from t he a mount of oxygen functional groupsand the hydrogen bond energies given above. Similarcorrelat ions were obtained for the heats of wett ing inwa ter and chlorobenzene a nd corresponding hydrogen

bond energies, where the hydrogen bond energies forOH.. .O a nd OH.. .Cl w ere, respectively, assumed t o be4.5 an d 3.9 kcal/mol. These pioneer works ar e veryinterest ing in the sense tha t t hey related the hydrogenbonding energy to a direct ly measurable qua ntity, hea to f we t t in g , a l t h o u g h i t ma y be q u e s t io n a ble i f t h e irinterpretation is correct.

One of the indirect m ethods for t he estima tion of heatassociated with a hydrogen bond is the est imation ofisosteric adsorption enthalpy (heat of adsorption) fromthe a dsorption isobars. Allardice and Eva ns est ima tedt h e h e a t o f a ds o rp t io n o f wa t e r du rin g t h e c o u rs e o fde sorp t ion of w a t e r f rom a ra w brown coa l .16 I t i n -creased from the level of heat of vaporization (43.88 kJ /mol-H 2O a t 25 ° C) t o 60 kJ /mol-H 2O with the decreaseof the amount of water adsorbed. This showed that coalhas different adsorption sites for water; in other words,coal oxygen functional groups have different hydrogenbon din g a bi li t ies . Wh e n t h e h e a t of a ds o rpt ion wa sest imated from the readsorption isobars for the driedcoal, on the other hand, its value was almost the sameas the hea t of vaporizat ion irrespective of the a mountof water adsorbed. This result shows that most of thehydrogen bonds formed among the oxygen functionalgroups in the dried coal are stronger than the hydrogenbonds th a t can be formed betw een th e oxygen functional

g r ou ps a n d w a t e r . Th is s h ow s t h a t t h e s t r e n gt h ofhydrogen bonds in the dried coal cannot be est imatedfrom w at er a dsorption dat a. To overcome this dra wba ck,G la s s a n d L a rs e n p res en t e d a me t h od t o s e pa ra t e t h eheats of adsorption of organic bases on coal into non-specific and specific interactions.17 The specific intera c-t ion heats, -q st,specific, which a re judged to be the inter-a c t ive e n erg ies bet we e n t h e ba s e s a n d coa l ox y g enfunctional groups, a re found to be equal t o the heat s ofh y drog en bon d f orma t io n be t we e n t h os e ba s es a n dp -fluorophenol in liquid phase,18 -∆H f , P F P , as shown inFigure 1. This clarif ied t ha t the ba se-coal interactiona n d ba s e-phenol interaction in the liquid phase can bet re a t e d in t h e s a me w a y . Th is a n a ly s is me t h od w ou ld

present a possibility to est ima te the st rength distribu-tion of hydrogen bonds in coal by changing the basessystematically.

Infrared (IR) spectroscopy is surely one of the mostpromising methods to qua ntify t he hydrogen bonds incoal directly.19-29 As P aint er et a l . reviewed,21 Fouriertransform infrared spectroscopy with on-line microcom-

(10) Larsen, J . W.; Gurevich, I. En ergy Fu els 1996, 1 0 , 1269-1272.(11) Lucht, L. M.; La rson, J . M.; P eppas, N. A. En ergy F uels 1987,

1 , 6-58.(12) Mackinnon, A. J .; H all , P. J . Energy Fuels 1995, 9 , 25-32.(13) Yun, Y.; Suuberg, E. M. Fuel 1993, 72 , 1245-1254.(14) Iyengar , M. S .; La hiri, A. Fuel 1957, 36 , 286-297.(15) Pa uling, L. The Nature of the Chemical Bond ; Cornell Univer-

sity P ress: Itha ca, NY, 1948; p 284.

(16) Allardice, D. J .; E van s, D . G . Fuel 1970, 50 , 236-253.(17) Glass, A. S.; La rsen, J . W. Energy Fuels 1994, 8 , 284-285.(18) Arnett, E. M.; Mitchel, E. J .; Murty, T. S. S. R.; Gorrie, T. M.;

Schleyer, P. V. R. J. Am. Chem. Soc. 1970, 92 , 2365-2377.(19) Solomon, P. R. Relation between Coal Structure and Thermal

Decomposition Products. In Coal Str ucture: Advances in Chemi stry Series ; Gorba ty, M. L., Ouchim, K., Eds.; ACS: Wash ington, DC, 1981;Vol. 192, Chapter 7, p 95.

Figure 1. Specific adsorption enthalpy, q st,specific, on deminer-alized Il l inois No. 6 coal vs heat of hydrogen bond formationw i t h p -fluorophenol, ∆H f , PF P , for seven bases. 17

600 E n er gy & F u el s, V ol . 15, N o. 3, 2001 M i u r a et al .

http://dontstartme.literatumonline.com/action/showImage?doi=10.1021/ef0001787&iName=master.img-000.png&w=225&h=147


3/12

puters has opened a new era of IR spectrometry. It hasbecome a rou t in e wo rk t o me a s u re I R s pe ct ra a n d t omanipulate them by spectral subtraction, curve resolv-ing, factor a na lysis, etc., wh ich lead the I R spectroscopyt o a q u a n t i t a t iv e a n a ly s is me t h od. H o we ve r , t h e re a reseveral problems to be solved to apply the FTIR methodto quantitat ive analysis of coal. The most commonlyemployed sample preparation method is the so-calledKBr pellet method. A small amount of coal (1 to 3 wt

%) is mixed wit h KB r powders, a nd t hey a re pelletizedt o a t ra n s p a re n t dis k t h a t is s e rv e d t o t ra n s mit t a n c eI R m e a s ur em en t . I t i s e ss en t i a l t o g r in d b ot h coa lpart icles and KBr part icles into the size less than thewavelength of the IR radiat ion. The most severe dif-ficulty is the ubiquitous presence of w at er. I t seems t obe a lmost impossible to remove wa ter effectively, espe-c ia l ly wh e n we in t e n d t o e s t ima t e t h e a mo u n t o f O Hgroups from the OH stretching absorption ba nds a round3000 to 3600 cm-1, s in ce s t ron g a n d b roa d w a t e radsorption bands appear in the region. I t is also verydifficult t o do subtra ctions of spectra on a 1/1 basis toobtain a difference spectrum without proper internalsta nda rds, becau se it is impossible to prepare tw o disks

under identical conditions.Overcoming these problems, Solomon et al. tried to

quantitat ively est imate the amounts of OH,23 a l ip h a t ichydrogen, an d ar omatic hydrogen 30 using the KB r pelletm et h od i n a s er i es of w o rk s w h e r e c a r ef ul s a m p lep re pa r a t i on a n d e la b or a t e d a t a a n a l y s is w e r e p er -formed. I t is reported t hat drying th e KB r pellet for 48h at 105 °C is enough to eliminate KBr-H 2O ba n ds a n dtha t t he absorptivity a t 3200 cm-1 correlates very wellwith the tota l a mount of OH irrespective of coal ra nk.O n t h e o t h e r h a n d, t h e a bs orp t iv i t ie s f or a l iph a t ichydrogen, a a l, a n d a roma t ic h y dro ge n , a a r, w ere foundt o be coa l ra n k de pe n den t . Th e ir v a lu es e st ima t e d,respectively, from the sets of peaks near 2900 cm-1 a n d

n e a r 800 cm-

1 we re a a l ) 7 4 6 a n d a a r ) 686 forbit u min ou s c oa ls a n d a a l ) 710 a n d a a r ) 541 forsubbituminous coa ls an d lignites in units of mg-1 cm-2.

So lomon e t a l . t r ie d t o e s t ima t e t h e a mou n t of O Hdirect ly f rom t h e O H s t re t ch re g ion t o s imp li fy t h ee st i m a t ion , P a i n t er e t a l . p rop os ed t o m ea s u r e t h es pe ct ru m ra n g in g f rom 1550 t o 1900 c m-1 of t h eacetylated coal for the est imation of OH amount.24 B ythis method acetylated phenolic OH, alkyl OH, and NHgroups were est imated from the bands of 1770, 1740,and 1670 cm-1, respectively. This m ethod w a s employed

to overcome a number of problems associated with theuse of the a bsorption band near 3400 cm-1 as a directmeasure of OH group content. The main problems arewa t e r a ds o rp t io n by t h e K B r ma t r ix , wa t e r bo u n d t ocoal in a complex manner, and the presence of morethan one band in this region of the spectrum. Thesefactors were judged to conspire to make quantitat ive

measurements based on the 3400 cm-1

band suspect .This w ork and the w ork performed by Sobkobiak et a l.28

showed tha t t he KB r pellet method is valid a nd precisewhen analyzing the bands except 3000 to 3600 cm -1.

There ar e severa l problems to qua ntify t he a bsorptionbands near 3400 cm-1 as st at ed above, but the absorp-tion bands ranging from 2400 to 3600 cm-1 containimporta nt informa tion on OH gr oup a ssociat ed hydro-gen bonds. Fuller and Smyrl25 showed tha t t he in-situdiffuse reflecta nce IR F ourier tra nsform (DRIF T) tech-n iq u e wit h n e a t , u n dilu t ed, c oa l s a mp les ca n be we llutilized to trace the in-situ reactions such as oxidation,dehydrat ion, etc.22 This technique seemed to resolvemost of the problems associated with the KBr pellet

method: no sloping baseline appears; wa ter a dsorptionby KB r does not occur; it is very easy to do subtra ctionsof spectra on a 1/1 basis to obta in a difference spectr um.Analyzing difference spectra obtained by this technique,F u ll er a n d S m y r l cl a r if ie d t h e a t t r i b ut i on s of t h eabsorption bands ranging from 2800 to 3625 cm-1 a sfollowss3625 cm-1: wea kly hydrogen bonded entit ies;3540, 3390, 3290 cm-1: structural difference and vari-ous types of hydrogen bonds; 2850-3000 cm-1: aliphatichydr ogen; 3000-3100 cm-1: a ro ma t ic h y drog en .

L a r s e n a n d B a s k a r r e l a t e d t h e s t r e n g t h o f c o a l-solvent interactions with the most direct spectroscopicp robe a v a i la ble : ch a n g e s in t h e h y drox y l s t re t ch in gfrequency on hydrogen bond forma tion.31 The spectrum

o f v a c u u m- drie d c o a l mix e d wit h K B r wa s me a s u re dusing a contr olled-at mosphere cell by th e DRI FT tech-nique. The coal w a s th en exposed to deutera ted solventv a por , a n d t h e s pe ct r u m of t h e coa l a d s or b in g t h esolvent wa s mea sured. The value of the OH stretchingf r eq u en cy s h if t w a s e st i m a t ed f r om t h e d if fe re ncespectrum of the tw o spectra. F igure 2 shows the w ave-n u m b e r s h i f t a g a i n s t t h e pK a v a lu e o f t h e s o lv e n t s(bases) used. The linear rela tionship wa s similar to tha tobtained when similar bases form hydrogen bonds tophenol in liquid phase. The good correlation in Figure2 s h ow s t h a t t h i s t e ch n iq u e h ol ds t h e p rom i se of

(20) Solomon, P . R.; H am blen, D. G .; Ca ra ngelo, R. M. Applicationof Fourier Transform IR Spectroscopy in Fuel Science. In Coal and Coal Products: Anal ytical Char acteri zation Technique; Symposium

Series ; F ul le r, E . L . , E d. ; AC S: Washi ng ton, DC , 1 9 82 ; Vol . 2 05 ,Chapter 4, p 77.(2 1) Pai nte r, P . C . ; Sny de r, R. W.; Sta rsi ni c, M.; C ole man, M. M.;

K u eh n , D . W. ; D a v i s, A. F o u r i er T r a n s f o r m I R Sp ec t r os co py sAp p l i c a t i on t o t h e Q u a n t i t a t i v e D et er m i n a t i o n of F u n c t i on a l G r o u p s ;Chapter 3, p 47, idem.

(22) Smy rl, N. R.; F uller, E. L ., J r. Chemistry and Structur eof Coals - Di ffuse Reflectance IR Fouri er Tr ansform (DRI FT) Spectroscopy of

Ai r O x i d a t i o n ; Chapter 5, p 133, idem .(23) Solomon, P . R.; C ar an gelo, R. M. Fuel 1982, 61 , 663-669.(24) Snyder, R. W.; Painter, P. C.; Havens, J . R.; Koeniug, J . L. Ap p l .

Spectrosc. 1983, 37 , 497-502.(25) Fuller, E. L., J r.; S myrl, N. Fuel 1984, 64 , 1143-1150.(26) Painter, P. C.; Sobkobiak, M.; Youtcheff, J . Fuel 1987, 6 6 , 973-

978.(27) Mu, R.; Malhotra , V. M. Fuel 1991, 70 , 1233-1234.(28) Sobkobiak, M.; Painter, P. C. Energy Fuels 1995, 9 , 359-363.(29) Chen, C .; G ao, J .; Yan, Y. Energy Fuels 1998, 12 , 446-449.(30) Solomon, P . R.; C ar an gelo, R. M. Fuel 1988, 67 , 949-959. (31) L a rs en , J . W.; B a s ker , A. J . Energy Fuels 1987, 1 , 230-232.

Figure 2. Plot of frequency shift of the O-H stretching bandof Rawhide coal vs base strength of the adsorbed solvent. 31

E st im at ion of H yd r ogen B on d D i st r i bu ti on i n Coal E ner gy & F uel s, V ol . 15, N o. 3, 2001 601



4/12

providing qua ntita tive informa tion of relative hydrogenbond st rengths.

P a in t e r e t a l .26 reviewed a wealth of l iterature con-cerning hydrogen bonds, and at tempted to define ex-plicitly the types of hydrogen-structure present. Theyclarified that the functional groups present in coals arecapable of forming hydrogen bonds with various accep-tor groups a re hydr oxyls, a nd phenolic and carboxylicacids, and a ssigned the OH stretching frequency ba nds

to the following hydrogen bondss

3611 cm-1

: f ree OHgroups;3516 cm-1: OH-π hydrogen bonds; 3400 cm-1:self-associated n -mers (n > 3);3300 cm-1: OH-ether Ohydr ogen bonds;3200 cm-1: t ig h t ly bou n d c y cl ic O Htetramers;2800-3100 cm-1; OH-N (acid/ba se str uc-tures).

Existences of the bands were determined by a curve-resolving method based on the second derivative of thes pe ct r u m a n d t h e ey e a n d b r a in e xa m i n a t i on . Th eOH-π hydrogen bonds are judged part ly to be due toOH groups that sterically cannot f ind a n energeticallypreferable partner. This implicitly meant that very few OH groups exist a s free OH. They a lso clarified tha t t hef r eq u en cy s h if t s f r om t h a t of t h e f r ee O H a n d t h e

in t en s i t ie s o f t h e ba n ds p rov ide f u n da me n t a l p a ra m-eters that can be used to determine structural charac-teristics such as bond lengths and the enthalpy of bondformation, although they did not est imate them.

As briefly reviewed, it has been clarified that thereare several types of hydrogen bonds in coal, and thatt h e ir e xis t en ce s a re mo s t v iv idly s h own by t h e O Hstretch frequency shift of IR absorption bands. However,no definite methods have been presented to est imatethe strength distribution of hydrogen bonds, probablybecause of the diff iculty in analyzing the OH stretchf req u e n cy ba n ds a s s o cia t e d w it h wa t e r bo u n d t o c oa lin a complex manner a nd t he presence of more tha n oneband in the region of the spectrum. I f the KBr pellet

me t h o d wa s e mp lo y e d, wa t e r a ds o rp t io n by t h e K B rma trix should a lso be taken into account.

Apart from coal hydrogen bonds there have been awea lth of works on va rious hy drogen bonds, includingvarious reviews and several books.32-34 In liquid-phasehydrogen-bonded adduct formation reactions betweenthe OH functiona l groups of phenols and /or a lcohols a ndvar ious ba ses have been exa mined in deta il using FTIRand thermodynamic considerations. The reaction for aphenol, for example, is represented by the followingstoichiometric equation:

wh e re P h-OH ‚‚‚B represents t he hydrogen-bonded ad-du ct , a n d ∆H i s t h e e n t h a lpy ch a n g e o f t h e re a ct ion .Th e ∆H va lues for va rious combinat ions of phenols a ndbases were est imat ed by applying the van’t H off equa-tion to the equilibrium constants measured. The amountof hydrogen bonds is also obtained in the course of themeasurement of equilibrium concentrations. When theh y dro g e n bo n d is f o rme d, t h e I R wa v e n u mbe r o f t h e

O-H stretching vibrat ion of the phenol shifts t o a low wa v e n u mbe r a n d g e n e ra t e s h e a t (∆H < 0). The rela-tionships between th e entha lpy chan ge, ∆H , and th e OHwa v en u mber s h i ft , ∆νOH , we re e x a min e d f o r v a r io u sphenol-base a nd a lcohol-base combinations18,35-39, anda l in e a r re la t io n wa s f o u n d t o h o ld be t we e n ∆H a n d∆νOH by ma n y in v e s t ig a t o rs . 38,40 The relat ionship iswell-known as “linear enthalpy-spectroscopic shift rela-t ions”. The relat ionship obtained by Dra go et a l . ,38 for

exam ple, from man y mea surements is given by

The values of -∆H w ere interpreted as the strength ofhydrogen bond by many investigators. 10-13 However,D ra g o e t a l . t h ou g h t t h a t ∆H consists of two contribu-tions: the decrease in the O-H bond energy due to theforma tion of the hydrogen-bonded adduct , δD OH (>0),a n d t h e bon d dis socia t ion e n erg y of h y drog en bon dformed, D H B , a s

They thought t ha t D H B i s a n in de x be t t er t h a n ∆H for

the strength of the hydrogen bond. Then a relationshipbetween δD OH a n d ∆νOH i s r eq u ir ed t o o b t a in t h erelationship betw een D HB a nd ∆νOH . They a pproxima tedt h e s t r et ch in g of t h e h y d r og en -b on d ed O H b y t h eanh ar monicity stretching of free OH, and related δD OHa nd ∆νOH theoretically as follows:

wh e re h is P lank’s consta nt , c is th e speed of light, D OH fis the bond dissociation energy of free O-H () 420kJ /mol), and νe is the w avenumber of harm onic vibra t iono f t h e f re e O H in c m-1, a n d e i s t h e a n h a rmo n ic i t y

constant for the oscillator.Using eqs 3 a nd 4, eq 2 can be solved t o obta in D HB

as follows:

N ow ei th er eq 2 or eq 5 d ir ect ly r el a t es t h e O Hwa venumber shift to the str ength of hydrogen bonds forvarious phenol-base a nd a lcohol-base hyd rogen bondsin liquid phase.

New Method Developed for E stimating StrengthDistribution of Hydrogen Bonds in Coal41,42

Definitions and Basic Concepts of the New

Method. We have at tempted to develop a method to

(32) Hydrogen Bonding ; Hadzi, D., Thompson, H. W., Eds.; Perga-mon Press Ltd.: London, 1959.

(33) Hydr ogen B onding ; Vinogradov, S. N., Linnell, R. H., Ed s.; VanNostran d Reinhold Co.: New York, 1971.

(34) T h e H y d r o gen Bo n d , R ec en t D ev el o p m en t s i n T h e or y a n d Experiments ; S c h u s t e r , P . , Z u n d e l , G . , S a n d o r f y , C . , E d s . ; N o r t h -Holland P ublishing Co.: New York, Oxford, 1976.

(35) J oesten, M. D.; Drago, R. S. J. Am. Chem. Soc. 1962, 8 4 , 3817-3821.

(36) Purcell , K. F.; D ra go, R. S. J . Am . Chem. Soc. 1967, 8 9 , 2874-2879.

(37) Epley, T. D.; D ra go, R. S . J. Am. Chem. Soc. 1967, 89 , 5770-5773.

(38) Drago, R. S.; E pley, T. D. J. Am. Chem. Soc. 1969, 91 , 2883-2890.

(39) Arnett, E. M.; Mitchel, E. J .; Murty, T. S. S. R . J . A m . C h em .Soc. 1974, 96 , 3875-3891.

(40) Sherry, A. D. Spectroscopic and Cal orim etr ic Stud ies of Hydr o- gen Bonding in Lit . 34.

(41) Miura , K.; Ma e, K.; Morozumi, F. Prepr. Pap.sAm . Chem. Soc.,Div. Fuel Chem. 1997, 42 (1), 209-213.

(42) Miura, K.; Mae, K.; Li, W.; Kusakawa, T.; Kumano, A. Prepr.Pap.sAm. Chem. Soc., Div. Fuel Chem. 1999, 44 (3), 642-646.

P h-OH + B f P h-OH ‚‚‚B ; ∆H (1)

-∆H ) 0.067 ∆νOH + 2.64[k J /m ol] (2)

∆H ) δD OH - D H B (3)

δD OH ) (h cN /4 e) ∆νOH ) D OH f (∆νOH /νe) )

0.131 ∆νOH (4)

D H B ) 0.198 ∆νOH + 2. 64 [k J /m ol ] (5)



5/12

q u a n t i fy t h e h y d rog en b on d s i n coa l b a s ed on t h efindings reviewed a bove. The hyd rogen bond forma tionreaction of OH groups can be defined by the followingequation:

where Coa l-OH represents a hypothetical sta te in w hichthe OH exists as a free OH group. The bases, B, in thiscase represent O a nd N in OH, COOH, CdO, and N-H

groups.We h a v e de cided t o a n a ly ze t h e O H s t re t ch in g a b-

sorption bands, because the informat ion on the hydro-gen bonds is concentrated on the absorption bands. Todo s o , s e ve ra l p roblems mu s t be s o lv ed. H ow do w eeliminat e the effect of adsorbed wa ter intera ct ing withcoal int imately? How do we determine the amount ofeach hydr ogen bond from t he complex absorption ba nd?Are eqs 2 and 5 applicable to the hydrogen bond in coal?The wa ter problem a ssociat ed with KB r w as overcomeby using the D RIFT technique w ith nea t coal samples.

The effect of adsorbed w at er w as not intended to beeliminat ed but intended to be included in the a na lysis,s in ce t h e a ds orbed wa t e r a f f ect s s ign ifica n t ly t h e h y -

drogen bonds in coal. Then the following hydrogen bondformation react ion was taken into account in addit ionto reaction 6.

wh e re H O-H again represents a hypothetical state inwh ich the OH exists a s a free OH group. The bas es, B ′,in this react ion involves O in water in addit ion to thebases in coal. This treatment looks apparently to makethe a nalysis complicated, but in rea lity this fa cilitatedt h e a n a ly s is a s s t a t e d below.

The amounts of individual hydrogen bonds, in otherwo rds t h e a mo u n t s of O H con t r ibut in g t o di f f ere n t

hydr ogen bonds, w ere determined by t he curve-resolvingmethod. First , the a bsorption ba nd r anging from 2400to 3750 cm-1 was regarded to consist of the following10 a bs o rp t io n ba n ds : 7 O H s t re t c h ba n ds a n d 3 CHstretch bands.

3640 cm-1; free OH groups3530 cm-1; OH-π hydrogen bonds3400 cm-1; self-associated n -mers (n > 3)3280 cm-1; OH-ether O hydrogen bonds3150 cm-1; t ightly bound cyclic OH tetra mers2940 cm-1; OH-N (a cid/ba se str uctur es)2640 cm-1; COOH dimers3050 cm-1; aromatic hydrogens2993, 2920 cm-1; a liphat ic hydrogens

The peak posit ions are a lit t le different from thoseproposed by Painter et al . ,26 and the hydrogen bondsa t t r i b ut a b l e t o C O OH d im er s w e r e a l so t a k en i n t oa c cou n t ju dg in g f rom t h e s econ d de riv a t iv e of t h espectrum and the brain a nd eye examina tion. Figure 3s c h e ma t ic a l ly s h o ws t h e h y dro g e n bo n ds t a k e n in t oaccount in this w ork. Assuming the G aussia n distribu-tion for each absorption band, the spectrum was curve-resolved to obtain t he intensit y of each ba nd. Of course,more complicated functions such as the Voigt function(mixture of Ga ussian and Lorentzia n functions) can beutilized for the curve resolving.21 However, we judgedthat the Gaussian distribution is precise enough for thecurve resolving as wa s suggested by S olomon et al . 20

Next, the a bsorptivity or t he extinction coefficient ofeach band must be determined to convert t he individualintensities to corresponding amounts of OH groups. Itis known that the absorption intensity of OH stretchfrequency forming a hydrogen bond increases w ith t heincrease of frequency shift, ∆νOH . Therefore, it is n eces-s a ry t o e s t ima t e t h e a bs o rp t iv i t y o f e a c h a bs o rp t io nban d independently in principle. Detoni et a l. exa minedthe change of absorptivity of the NH stretching band ofisothiocyanic acid (HNCS ) a ccompanying the forma tionof h y d rog en b on d s w i t h v a r iou s a c ce pt or s i n C C l4systematically.43 They found t he follow ing rela tionshipbetween t he a bsorptivity of t he hydr ogen-bonded NH ,R

NH , a n d t h a t of f ree N H , R

NH,0:

where ∆νNH is the NH wa venumber shift. This relation-ship shows that the absorption intensity of the hydrogen-bonded NH increases in proport ion to ∆νNH . O n t h eb a s i s o f t h i s w o r k w e a r b i t r a r i l y a s s u m e d a s i m i l a rrelat ionship for the change in absorption intensity ofhydrogen-bonded OH , a nd est imated the proport ionalconstant from the intensity and the ∆νOH values mea-sured for several model compounds (benzoic acid and2-na phthol) a s

wh e re R OH is the absorptivity of the hydrogen-bondedO H a n d R OH,0 is the absorptivity of the free OH. Theproportionality consta nt obta ined for t he OH-associat edhydrogen bonds (eq 9) wa s found to be very close to tha tfor the NH-associated hydrogen bonds (eq 8). Thus wecan calculate the amounts of OH contributing to differ-ent hydrogen bonds using eq 9 when we can est imateR OH,0.

The last problem to be solved is whether eqs 2 and 5a re a p p lica ble t o t h e h y drog e n bon d in coa l . Th e ir

(43) Detoni, S.; Hadzi, D.; Smerkolj, R. J. Chem. Soc. (A) 1970,2851-2859.

Co al-OH + B f Co al-OH ‚‚‚B ; ∆H (6)

H O-H + B ′f H O-H ‚‚‚B ′; ∆H ′ (7)

Figure 3. Schematic representation of hydrogen bonds as-sumed to exist in coal in this work.

R NH ) R N H, 0(1 + 0.0141 ∆νNH ) (8)

R OH

) R OH,0

(1 + 0.0147 ∆νOH

) (9)




6/12

applicability is supported by the works of Larsen et al.(Figures 1 and 2). Figure 2 qua litat ively shows tha t theO H s h if t in c re a s e s wit h ba s e s t re n g t h a s in t h e c a s efor phenol-base hyd rogen bonds, and Figure 1 suggeststhe equiva lence of the ∆H () q st, specific) values a nd ∆H f,P FP va lues. These results suggest tha t t he coal h ydrogenbonds ar e treated in a similar w ay a s the phenol-ba s ehydrogen bonds. Although not direct ly related to thecoal hydr ogen bonds, K iselev44 found a linear relat ion-

ship between the heat of adsorption, ∆Q a , a n d t h e ∆νOHf or a h y drox y la t e d s i lica s u rf a ce a n d v a riou s v a p orsincluding steam. I t was represented by

These works clarif ied that the OH wavenumber shiftcaused by the formation of hydrogen bonding is wellrelated to the enthalpy change of the hydrogen bondingf or m a t i on . B a s ed on t h es e con s id er a t i on s w e h a v ejudged tha t eqs 2 a nd 5 a re a pplicable to the hydr ogenbond in coal a s a f irst a pproximat ion. Using eqs 2 a nd5, w e ca n n ow e st i m a t e t h e s t r e n gt h of i nd i vi du a lhydrogen bonds in coal from the ∆νOH va lues.

Procedure To Estimate the Strength Distribu-tion of Hydrogen Bonds. The procedure to obtain thestrength distribution of hydrogen bonds by the proposedmethod is summar ized as follows:

1. Divide the FTIR spectrum ranging from 2200 to3700 cm-1 into 6 hydrogen-bonded OH bands, 3 CHstretch bands, a nd the free OH ban d, if i t existed, by acurve fit t ing method as schematically given in Figure4a . T h e G a u s s ia n dis t r ibu t io n wa s a s s u me d f o r e a c habsorption ba nd in t his work. Three CH stretch bandsa re e a s i ly e limin a t e d f rom t h e s p e ct ru m. S in ce w eemployed in situ mea surements, three CH st retch bandsdid not change during the measurement. It m ade it easyto strip off the CH bands from the OH bands.

2. Estimate the amount of OH for the j th peak, (n OH ) j ,by the Beer law as follows:

wh e re A j a n d R j are, respectively, the integral intensityand the a bsorptivity of the j th peak. In this analysis R j was assumed to be represented by eq 9.

wh e re R 0 is the absorption coefficient of the stretchvibrat ion of the free OH. Then the amount of the OH

corresponding to the j th peak, (n OH ) j , is given by

3. Ca lc ula t e t h e s t re n g t h o f t h e j th hydrogen bond,(-∆H ) j by e q 2 a s

The value of (D H B )j c a n a ls o be o bt a in e d f ro m e q 5 i fnecessary.

4. (n OH ) j v s (-∆H ) j relat ionship gives the distribut ionof the strength of hydrogen bonds a s shown schemati-cally in Figure 4c.

Experimental Section

Samples. Seven Argonne premium coals and an Australianbrown coal (Morwell) were used a s coal sa mples. The ultima teanal y se s and t he am ount of O H and C O O H oxy g e n 45,46 a r egiven in Table 1. Coal samples -200 mesh were furth er ground

for 1 m i n i n a m or t ar just b efor e m e asur e m ent . As m ode lcompounds for coal a n ion-excha nge resin (Mitsubishi C hemi-cals, WK11) that has carboxylic groups as ion-exchange sites(Figure 5) and a l ignin (Naka lai Tesque, Co.) were a lso used.The amount of COOH in the resin was estimated to be 10.3mol /kg by 13C NMR measurement.

Apparatus and Procedure.The d iffuse reflecta nce spec-t r um at any sam pl e , R , w as m e asur e d at 4 c m-1 resolutionson a J E O L J I R -WI NS P E C 50 FTI R spec t r om e t er w i t h amicroscope stage. A Mettler 84HPT hot stage, shown sche-

(44) Kiselev, A. V. Su r f . Sc i . 1965, 3 , 292-293.

(4 5) J ung , B . ; St ac he l, S . ; C a l kins, W. H. Pre pr. P apsAm . C h em .Soc., Di v. Fu el Chem. 1991, 36 (3), 869-876.

(46) Schafer, H . N. S. Fun ctional Groups and I on Exchange Proper- ties. In T h e Sci en c e of Vi c t or i a n Br o w n C oa l ; D u r i e , R . A . , E d . ;Butt e rworth: He i nemann, 1 99 1; C ha pte r 7 , p 3 34 .

∆Q a ) 0.064 ∆νOH - 1.16 [kJ /mol] (10)

(n OH ) j ) (1/R j )A j (11)

R j ) R 0{1 + 0.0147(∆νOH ) j } (12)

(n OH ) j ) 1

R 0{1 + 0.0147(∆νOH ) j } A j (13)

(-∆H ) j ) 0.067 (∆νOH ) j + 2.64 [kJ /mol] (2)

Figure 4. Procedure to obtain the strength distribution ofhydrogen bonds by the proposed meth od.




7/12

m at i c al ly i n Fi g ur e 6, w a s at t ac he d t o t he m i cr osc ope st ag e

for the in-situ measurement. About 0.5 mg of neat coal samplesw e r e l e ve l e d and pr e sse d on t he m i r r or b y hand usi ng aspat ula. It is essentia l to make a flat s urfa ce for obta ining goodand reproducible spectra. The coal samples were equilibratedat 30 ° C in a flowing nitrogen stream . Then they w ere heatedat t he r a t e of 5 K /m i n up t o 300 ° C d ur ing w hi ch ul t r ahi g hpur it y ni t r og e n w a s cont inual l y suppl ie d at t he r a t e of 100cm 3/m in t o p re ve nt a n y p os s ib le l ea k a g e o f a i r i n t o t h echamber and to keep the sta bili ty of the tempera ture. A Ba F 2

d is k w a s p ut on t h e p in h ol e f r om w h e r e t h e l ig h t p a s se s

through. The l ight source was a heated Ni-C r w i r e a n d t h e

detector was MCT (mercury cadmium telluride). The diffuse

reflecta nce spectru m w as collected by a cquisition of 100 scans

at 20 °C intervals. A mirror, which is believed not to absorb

water, was used to obtain the background diffuse reflectancespectrum, R 0. The ra tio of R /R 0 wa s converted to the Ku belka-

Munk (KM) function to obtain the spectrum corresponding to

the a bsorban ce. The spectrum t hus obtained a t a ny sa mple is

simply called an FTIR spectrum in this pa per.

Results and Discussion

Validity of in-Situ DRI FT Technique Employed.As discu s s ed a bo ve , t h e re a re s ev era l e xp erimen t a lproblems to be solved for analyzing an FTIR spectrumqua ntit a tively. However, most of the problems could besolved by using th e in-situ D RIF T technique. The wa terad sorption problem w as a lso overcome by a na lyzing thespectrum measured with the adsorbed water. Figure 7shows t he change of the FTIR spectra for th e resin with

Table 1. Ultimate Analyses and Oxygen Distribution of 8 Coals Used [wt %]

ult im a t e a n a ly sis [da f] oxy gen cont en t [da f]

coa l (a bbr ev.) C H N O(diff.) a s C OOH a s OH a s CdO a s et h er

Mor w ell (MW) 67.1 4.8 0.7 27.4 7.97 5.92B eula h -Za p (ND ) 72.9 4.8 1.2 20.1 3.81 9.16 1.96 2.28Wyoda k (WY) 75.0 5.4 1.1 21.0 3.33 7.68 0.74 2.56I llin ois #6 (I L ) 77.7 5.0 1.4 13.5 0.23 5.86 0.93 0.57B lin d C a n yon (U T) 80.7 5.8 1.4 13.5 0.23 5.22 0.63 3.68P it t sbur gh No. 8 (P I TT) 83.2 5.3 1.6 8.8 0.16 4.32 0.18 2.08U pper F r eepor t (U F ) 85.5 4.7 1.6 7.5 0.14 1.96 0.44 0.36

P oca hon t a s (P OC ) 91.1 4.4 1.3 2.5 0.05 0.98 0.32 0.24

Figure5. S tructu re of an ion-excha nge resin used as a m odelcompound for coal.

Figure 6. Schematic of a modified Metler 84HPT hot sta ge used for in-situ FTIR measurements.


http://dontstartme.literatumonline.com/action/showImage?doi=10.1021/ef0001787&iName=master.img-005.png&w=413&h=339http://dontstartme.literatumonline.com/action/showImage?doi=10.1021/ef0001787&iName=master.img-004.png&w=197&h=92


8/12

t h e in cre a s e o f t e mpe ra t u re . O n e of t h e p roblemse n cou n t e red by ma n y in ve s t ig a t o rs in u s in g in -s i t uFTIR w as t he spectra qua lity at high temperat ures due

to the cha nge in refra ctive index. Very beaut iful spectrawith a lmost f lat baselines show tha t such a problem iswell overcome by the D RIF T technique employed here.To e x a min e t h e ch a n g e of t h e O H a bs orp t ion ba n dregion, the spectra ra nging from 2400 to 3650 cm-1 wereenlarged a nd shown in F igure 8. The weight cha nge andt h e g a s f or m a t i on r a t e s m e a s u r ed u n de r t h e s a m eexperimental conditions using a TG-MS method are alsoshown in Figure 8 to clarify the change during the hea t-up of the resin. When the resin wa s hea ted from roomtempera ture t o 140 ˚C, t he intensity of t he a bsorptionban ds r a nging from only 2800 to 3400 cm-1 decreased.This decrease can be well associated with the desorptiono f a d s o r b e d w a t e r r e f e r r i n g t o t h e w a t e r f o r m a t i o n

profile in the lower figure. This also indicates that wateris adsorbing with OH groups by ra ther w eak hydrogenbonds. B y furth er heat ing, all OH-associa ted a bsorptionbands decrease significantly, and these decreases ac-company again the formation of water as shown in thelo we r f ig u re . T h e wa t e r f o rme d in t h is t e mp e ra t u rerange is judged to come from the decomposition reac-tions of carboxylic groups represented by

This can be confirmed by the appearance of dist inctabsorption ban ds of an a nhydride (-CO O CO-) a t n e a r1800 cm-1 as shown in Figure 7. Thus it was found that

the changes of the OH absorption bands measured bythe in-situ D RIFT technique give a lot of informa tion,including the react ion of the early stage of pyrolysis.Th is wil l be s h o wn more dis t in ct iv ely t h ro ug h t h eq u a n t i t a t iv e a n a ly s is ma de below.

Change of OH Absorption Bands of Coal withHeating. F ig u re 9 s h ows t h e in -s i t u F TI R s p ect rara ngin g fr om 2200 to 3700 cm-1 measured for the eightcoals. Very bea utiful a nd r eproducible spectr a could beobtained by use of the proposed measurement techniquefor all the coals. Very f lat baselines were obtained atall the temperatures, and the spectra for POC did notchange significantly above 70 °C. This suggested thatthe spectra were not affected by the temperature below

300 °C. Separate experiments were performed usingkaolinite to check the cha nge of spectra wit h increasingtempera ture up t o 350 ° C. The mirror ba ckgrounds at30 °C and at corresponding temperatures were used,respectively, to get the FTIR spectra. The spectra didnot change with the background at different tempera-t u re s . T h e n a l l t h e s p e c t ra s h o wn in t h is wo rk we reobta ined using the background a t 30 °C . This ma de thein-situ m easurement convenient , fast , and reliable.

To get a ccura te informa tion of hydr ogen-bonded OHgroups, it is essential to ensure that the art ifact doesnot exist during experiments and/or during the dat aprocessing after experiments. I t is easily checked by

tra cing the changes in the alipha tic and a romatic C-Hbands. Because the C-H bond is hard to decompose orto form below 300 °C , we can ea sily know wh ether someartifact exists or not from the variation of the 2890 and2930 cm-1 ba n ds (t h e a bs orp t ion of a l iph a t ic C-Hbon ds ), a n d t h e 3050 c m-1 ba n d (t h e a bs orp t ion ofaromatic C-H bond). It w a s found tha t th ese intensitieswe re n o t a f f e c t e d by t h e t e mp e ra t u re u p t o 300 ° C.Therefore, all th e changes in t he OH spectra were rea llywh a t h a p p en e d in c oa l du rin g t h e h e a t in g . N o w, s ig -nifican t cha nges in t he OH spectra were judged to occurfor MW, ND, and WY coals, whereas very little changewa s judge to occur for P ITT and POC coals during th eheating.

Figure 7. Change in the in-situ FTIR spectrum of the ion-excha nge resin (WK11) with the increase of temperature.

2-COOH f -COOCO- + H 2O (14)

Figure 8. C hang e s i n t he O H st r e t chi ng b and s, w e i ght , a ndgas formation rates during a heating of the ion-exchange resin(WK11).




9/12

Change in Strength Distribution of OH Associ-ated with Hydrogen Bonds. All the spectra measuredfor t he eight coals were a na lyzed by th e procedure givena b ov e t o e s t im a t e t h e c h a n g e i n H B D t h r ou g h t h eheat ing. Figure 10 shows the result of peak division a ndthe HBD obtained for ND as an example. The spectraranging from 2200 to 3650 cm-1 were divided into 9p ea k s by a cu rv e-f it t in g me t h od. Th e s ix t y p e s ofh y drog en bon ds w e re a bbrev ia t e d a s f ol lows : H B 1:OH-π at 3530 cm-1, H B 2: OH-O H a t 3400 cm-1,H B 3 : O H-ether at 3280 cm-1, H B 4: cy cl ic O H-OHat 3150 cm-1, H B 5 a t 2940 c m-1; OH-N , a n d H B 6 :COOH-COOH at 2650 cm-1. These a re in t he order of

increasing strength. No free OH bands were detectedf o r a l l t h e c o a l s . I n a c t u a l p e a k d i v i s i o n , t h e p e a kp os i t ion s we re s l ig h t ly ch a n g e d, de pe n din g on t h esample by referring to the second derivat ive of eachspectru m. The ba nds for t he C-H bonds did not changedu rin g t h e h e a t in g , t h e n t h e y we re e a s i ly s e p a ra t e dfrom the hydrogen-bonded bands. The amount of OHa n d t h e s t re n g t h of e a c h h y dro g en bon d, (n OH ) j , we recalculated from each peak intensity, A j , a n d t h e O H

wavenumber shift , ∆νOH , with the aid of eq 13. To doso, the R 0 value must be est imated for each sam ple. I fwe a s s u me t h a t a l l a ds orbed w a t e r is de s orbed be low 150 °C (the va lidity of this a ssumption is shown below),t h e t o t a l a m ou n t o f O H (t h e s u m o f O H a n d C O OHgroups) given for each coal in Table 1 is equated to thetotal a mount of OH a t 150 ° C. The total a mount of OH,(n OH )t o t a l, can be related to the intensity of each absorp-tion band by

By applying this relat ionship to the spectrum obtaineda t 1 5 0 ° C R 0 va lu e c a n be e s t ima t e d f o r e a c h coa lsample. The st rength distributions of hydrogen bondsthus est imated, (n OH ) j vs (-∆H ) j relationships, for NDare shown in the right figure in Figure 10. The strengthsof in div idu a l h y drog en bon ds re pre s en t e d by -∆H values a re HB 1: 10 kJ /mol-OH, HB 2: 17 kJ /mol-OH,H B 3: 25 kJ /mol-OH , HB 4: 35 kJ /mol-OH , HB 5: 50 kJ /mol-OH, H B 6: 70 kJ /mol-OH, a nd t hose represented byD H B va lues a re H B 1: 24 kJ /mol-OH, H B 2: 50 kJ /mol-OH, H B 3: 74 kJ /mol-OH, HB 4: 100 kJ /mol-OH , H B 5:141 kJ /mol-OH, HB 6: 199 kJ /mol-OH. Since we a s-sumed that all a dsorbed wa ter wa s desorbed by 150 ° C,

the distribution at 150 °C is the distribution of the coaldried and heated to 150 °C. The dried ND coal is richin ra th er wea k hydr ogen bonds: 2.5 mol-OH/kg of HB 1,1.3 mol-OH/kg of H B 2, a nd 1.1 mol-OH/kg of H B 3.Wh e n t h e ra w N D c oa l wa s h e a t e d f ro m 30 ° C t o 150° C, the a mounts of weaker hydrogen bonds such as HB 1,HB 2, HB3, and HB 4 decreased significantly. Since thisdecrease is d ue to the desorption of adsorbed w at er, i tis f ou n d t h a t wa t e r is in t e ra c t in g wit h coa l h y drox y lsby rather weak interactions. When the coal was furtherheated to 270 °C, all hydrogen bonds decreased. Thisde cre a s e is ju dg ed t o be du e t o t h e de comp os it ionrea ctions of carboxylic groups such a s eq 14. Thus it w a scla r i f ie d t h a t t h e in -s i t u D R I F T t ech n iq u e a n d t h e

an alysis m ethod proposed here are powerful to examinethe change in hydrogen bonds in coal.

Sensitivity of FTI R to Adsorbed Water. I n t h ecourse of various spectra measurements we have expe-rienced how FTIR is sensit ive to water. When a coals a mp le wa s h e a t e d t o 130 ° C t o re mov e t h e a ds orbedwa t e r a n d t h e n c o o le d t o 30 ° C in a n a t mo s p h e re o fflowing dry n itrogen, the a bsorption ba nds ra nging 3000to 3500 cm-1 in cre a s ed con s ide ra bly. We h a v e h a ddifficulty in interpreting the unexpected change, andfinally we found tha t it w as t he result of the adsorptionof water from the “dry” nitrogen. The dried coal tookup a fa irly large a mount of wat er very quickly even fromthe dry n itrogen containing w at er by less tha n 100 ppm.

Figure 9. I n-situ FTIR spectra measured a t every 20 °C forseven Argonne premium coals a nd a n Australian brown coal .

Figure 10. Six hydrogen-bonded OH stretching bands andthree CH bands obta ined by a curve resolving method a nd thest r e ng t h d i st r i b ut i ons of hy d r og e n b ond s e st i m at e d b y t heproposed method for ND coal.

n O H, t o t a l )∑ j

(n OH ) j )1

R 0

∑ j [

1

{1 + 0.0147(∆νOH ) j }A j ]

(15)




10/12

When w e used an ultra high purity nitrogen conta iningwater less than 10 ppm instead of the dry nitrogen asthe carrier ga s, very little change w a s observed betw eent h e s p e c t ru m a t 130 ° C a n d t h e s p e c t ru m o f t h e c o a lcooled to 30 °C a s typically show n for ND coal in Figur e11. The slight difference betw een the tw o spectra isproba bly due to rearr an gement of hydrogen bonds uponcooling. Thus we found that in-situ measurements in a

wa ter-free atmosphere a re indispensable for qua ntita -t iv e a n a ly s is o f O H a bs o rp t io n ba n ds . W e we re a ls oimpressed tha t the in-situ D RIFT technique w ould bea m os t s en s it i ve m et h od t o a n a l y z e w a t e r i n s ol idmaterials. These findings also suggested that the waterproblem cannot completely be removed when the con-ventional KBr pellet method was used.

Quantitative Analysis of Change in HydrogenBonds during the Heating. Th e ch a n g es of t h eamounts of the six hydrogen-bonded OH (HB1 to HB6)were estima ted by th e procedure presented a bove, thenthe changes of their amounts with increasing temper-ature are shown for the all eight coals in Figure 12. Forlower rank coals, MW, ND, and WY, the amounts of

we a k e r h y drog en bon ds s u ch a s H B 1, H B 2, H B 3, a n dHB4 decreased significantly below 150 °C. These de-creases a re a ssociat ed w ith t he desorption of adsorbedwa ter a s shown for the resin in Figure 8. With furtherincrease of tempera ture, the t otal a mount of hydrogenbonds, (n OH )t o t a l, decreased for MW, ND, and WY, but itwas kept almost constant for IL, PITT, UF, and POC.These results show tha t decomposit ion react ions ofcarboxylic groups star t a t a round 200 °C for th e lowerra n k coa ls w h ich h a v e ma n y CO O H g ro u ps a s s h o wni n Ta b l e 1 . O n t h e ot h er h a n d , s u ch r ea c t ion s a r enegligible for the higher ra nk coa ls, because the C OOHcontents ar e very low for these coals. I t is noteworthythat only the distribution changes without affecting the

total amount of hydrogen bonds for UT, PITT, and UFabove 200 °C. This means that the stronger hydrogenbond, HB5, shifted to weaker hydrogen bonds, HB1 andHB2 at higher temperatures. This phenomenon seemsto be associated with the glass transit ion.

To examine the change in the hydrogen bonds in moreq u a n t i t a t iv ely , t o t a l e n t h a lpy f or t h e f o rma t ion of a l lOH a ssociat ed hydrogen bonds, (- ∆H )t o t a l (


11/12

10 kJ /kg-coal a nd 7 kJ /mol-OH for U F, a nd 6 kJ /kg-

coal a nd 3 kJ /mol-OH f or P ITT.

Wh a t we w ou ld l ik e t o s t re ss h e re is t h a t t h e v a lu e

of R 0 is not required to calculate (-∆H )a v a s can be foundfrom eq 18, which means that FTIR spectra measure-

ments are self-sufficient to obtain (-∆H )a v. This is one

of the merits of the proposed analysis method. This

facilitates the analysis, and therefore the accuracy of

t h e (-∆H )a v values is well expected.

Differentiation between Adsorbed Water andDecomposed Water. I t h a s be e n ra t h e r di f f ic u l t t odistinguish the chemisorbed wa ter a nd the wa ter formed

by decomposit ion react ion, which made it diff icult to

define the dried sta te of coal especially for lower ra nk

Figure 12. C hang e of (n OH ) j , (n OH )t o t al, a n d (-∆H ) t o t al with increasing temperature for the eight coals.




12/12

coals. We arbitrarily assumed above that all adsorbedw a t e r d e s o r b s b y 1 5 0 ° C . N o w w e w i l l e x a m i n e t h evalidity of the a ssumption.

In th is ana lysis we could est imate the changes in thea mo u n t o f O H , (n OH )t o t a l, a n d t h e e n t h a lp y re la t e d t ohydrogen bonds, (-∆H )total, as s hown in Figur e 12. Thenwe c a n e st ima t e t h e ch a n g e in t h e e n t h a lp y re la t e d t ohydrogen bonding with the formation of wat er by usingthe values of (-∆H )t o t a l a n d (n OH )t o t a l a s

where 2 in the rh s w as included to a ccount for t he factt h a t o n e H 2O molecule is involved in two hydrogenbonds. The magnitude of (∆H )H 2O thus est imat ed is wellexpected to be dependent of the mechanism of water

format ion. Then t he (∆H )H 2O values calculat ed by eq 19w i t h t h e a i d o f e q s 2 , 13, a n d 1 6 a r e s h ow n a g a i n s ttemperature in Figure 14. The values obtained for theresin and the lignin a re a lso included in t he figure forcomparison. The (∆H )H 2O vs t empera ture relat ionshipswe re s o c los e t o e a c h ot h e r , a n d t h e re wa s a dis t in ctju m p a t a b ou t 1 40 ° C : t h e (∆H )H 2O values below 140° C we re c lo s e t o o r s l ig h t ly s ma lle r t h a n t h e h e a t o fv a p oriza t ion of wa t e r , ∆H va p, w h e r e a s t h e (∆H )H 2Ov a lu es a bo ve 140 ° C we re s ig n if ica n t ly la rg e r t h a n∆H va p. These results suggest that the water detectedbe lo w 140 ° C is t h e a ds o rbe d wa t e r a n d t h a t f o rme da bove 140 °C is the w a ter formed by some decompositionre a ct ion . Th u s t h e H B D e s t ima t e d by t h e p rop os e d

method is found to be utilized to disting uish th e type ofwa t e r f orme d. We s t re s s a g a in t h a t t h e v a lu e o f R 0 orthe va lue of (n OH )t o t a l is not required to obta in (∆H )H 2O,a s w a s t h e c a se t o c a lcu la t e (-∆H )a v. FTIR spectrameasurements alone are enough to obtain (∆H )H 2O.

Conclusion

A new method wa s presented to estimate th e strengthdistribution of hydrogen bonds in coal, a nd its va lidity

wa s exam ined by ana lyzing the chan ge in the hydrogenbond distributions of 8 different coals, a resin, and alignin during their heat treatment below 300 °C. Thehydrogen bonds examined included th e coal intra mol-ecule hydrogen bonds and coal-water hydrogen bondsformed by hydroxyls in coal. The method a na lyzes theFTIR spectrum ranging from 2400 to 3700 cm-1 t h a tare obtained using the in-situ diffuse reflectance IRFourier t ran sform (DRIF T) technique w ith neat , undi-luted, coal samples.

The ana lysis method and t he experimental t echniqueemployed clarified the following.

1. Wa t er i s t a k en u p o n coa l ev en f rom a d r yatmosphere so quickly that in-situ measurements in

completely wa ter-free at mosphere a re indispensable forq u a n t i t a t iv e a n a ly s is of O H a bs orp t ion ba n ds . Th issuggested that the water problem cannot completely beremoved when the conventional KBr pellet method wasused.

2. Coal itself ha s hydr ogen bonds ra nging from 10 kJ /mol to 70 kJ /mol in strengt h. The r elat ive a bunda nceof the hydr ogen bonds is high ly dependent on coa l type.Lower rank coals are rich in weaker hydrogen bondsassociated with OH-π , OH-H, and OH- ether, whereassome high rank coals are rich in OH-N hydrogen bonds.

3. The a verage st rength of hydrogen bonds w a s foundto decrease above 230 °C for some bituminous coalswithout a ffect ing the tota l number of hydrogen bonds.This clarif ied t hat the change accompanying the gla sstransit ion can be well traced by the proposed analysismethod.

4. Th e e n t h a lpy a s s ocia t e d wit h t h e f orma t ion ofwater could be successfully est imated from solely theFTIR measurement. The entha lpy increased gra duallyup to 140 ° C from 30 kJ /mol to 40 kJ /mol, but itincreased in a stepwise m a nner to a level of 50 kJ /molo v e r 140 ° C. T h is c la r i f ie d t h a t t h e wa t e r f o rma t io nbelow 140 °C is due to th e desorption of adsorbed wa ter,a n d t h a t t h e wa t e r f o rma t io n a bo v e 140 ° C is du e t othe decomposit ion react ions of C OOH groups. Thisf in din g e n a ble d u s t o de f in e t h e drie d s t a t e o f c o a ldefinitely. I t is a lso possible to exam ine qua ntita t ivelythe ear ly sta ge of pyrolysis of coal on the ba sis of thisinformation.

Th u s t h e v a l idi t y a n d u s ef u ln e ss of t h e p rop os edmethod was clarified. We are sure that this method willbe extended to analyze various solvent-coal interac-tions. The extension of this method to other fields mayalso be well expected.

Acknowledgment. Th is w ork wa s s u pp ort e d by“Research for the Future” project of the J apan Societyfor the P romotion of S cience (J SP S) th rough th e 148committee on coal utilization technology of J SPS.

EF0001787

Figure 13. C h a n g e i n t h e a v e r a g e s t r e n g t h o f h y d r o g e nbonds, (-∆H )a v, w i t h i nc r e asi ng t e m pe r at ur e for t he e i g htcoals.

Figure 14. Changes in the water formation enthalpy, (∆H )H 2O,with increasing temperature for the coals, the resin, and thelignin.

(∆H )H 2O ) 2d(-∆H )t o t a l

d (n OH )t o t a l[kJ /m ol-H2O] (19)


Estimation of Hydrogen Bond Distribution in Coal

Documents

Transcript of Estimation of Hydrogen Bond Distribution in Coal