Chapter III:電磁場的相干性質,一些基本概念

上一章講了一些電磁場量子化的基本概念,這一章講一些具體一點的東西。首先我們介紹關聯函數(correlation function)特別是一階和二階的。以及光學相干(opticalcoherence),壓縮真空,光子反聚束(photonantibunching)以及光子計數(photoncounting)相關的知識。

目錄如下:

Coherence Properties of the Electromagnetic Field 3.1 Field-Correlation Functions 3.2 Properties of the Correlation Functions 3.3 Correlation Functions and Optical Coherence3.4 First-Order Optical Coherence 3.5 Coherent Field3.6 Photon CorrelationMeasurements 3.7 QuantumMechanical Fields3.7.1 Squeezed State 3.7.2 Squeezed Vacuum 3.8 Phase-DependentCorrelation Functions3.9 Photon CountingMeasurements3.9.1 Classical Theory 3.9.2 Constant Intensity3.9.3 Fluctuating Intensity–Short-TimeLimit3.10 Quantum Mechanical Photon Count Distribution .3.10.1 Coherent Light 3.10.2 Chaotic Light 3.10.3 Photo-Electron Current Fluctuations

  • 關聯函數

當我們需要探測一個電磁場的時候,我們有一個辦法就是去數有多少個光子被吸收了。但是這種對於自發輻射是很不敏感的,這涉及到了光和原子之間的作用(mathbf{QED}),我們之後會講。這裡,我們考慮到一種理想的探測器對在(r,t)的電場E^{+}(r,t)非常的敏感。那麼這個透射概率(transitionprobability)我們可以這麼去描述:

T_{if}=| langle f|E^{+}(r,t)|i
angle|^{2},其中i,j代表初態和末態,所以為了計量所有的情況,我們需要遍歷整個過程,也就是把這個T_{if}加起來得到:I(r,t)=sum_{f}T_{fi}=sum_{f}langle i|E^{-}(r,t)|f
anglelangle f|E^{+}(r,t)|i
angle=langle i|E^{-}(r,t)E^{+}(r,t)|i
angle,當然末態是完備的sum_{f}|f
anglelangle f|=1.

以上說的都是純態(pure state),如果是混態(mixed state)只需要加一項概率項進去,得到:I(r,t)=sum_{i}P_{i}langle i|E^{-}(r,t)E^{+}(r,t)|i
angle,我們可以用跡(trace)來表示:

I(r,t)=Tr(
ho E^{-}(r,t)E^{+}(r,t)),其中
ho=sum_{i}P_{i}|i
anglelangle i|是密度矩陣算符(densityoperation),如果初態是在真空態下,有
ho=|0
anglelangle 0||.所以有I(r,t)=langle 0|E^{-}(r,t)E^{+}(r,t)|0
angle=0.

所以我們可以定義關聯函數了:mathcal{G}^{1}(x,x)=Tr{(
ho E^{-}(x)E^{+}(x)}),很自然的高階的關聯函數也被定義出來了:

mathcal{G}^{(n)}(x_{1}cdots x_{n},x_{n+1}cdots x_{2n})=Tr{{
ho E^{(-)}(x_{1}cdots E^{(-)}(x_{n})	imes E^{(+)}(x_{n+1})cdots E^{(+)}(x_{2n})}

  • 關聯函數的性質

線代里有對於非負項mathcal{A},有Tr{
homathcal{A}^{dagger}mathcal{A}}geq0,所以對於mathcal{A}=E^{(+)}(x),有關聯函數mathcal{G}^{1}(x,x)geq0.如果我們選擇mathcal{A}=sum^{n}_{j=1}lambda_{j}E^{(+)}(x_{j}),我們可以得到:

sum_{ij}lambda_{i}^{*}lambda_{j}mathcal{G}^{(1)}(x_{i},x_{j})geq0,係數是正定二次型的,所以這個關聯矩陣行列式值是正的。det[mathcal{G}^{(1)}(x_{i},x_{j})]geq0

我們用到(Cauchy-Schwarzinequality):對於向量u,v,有|langle u,v
angle|^2leqlangle u,u
anglelangle v,v
angle,其中langle cdot,cdot
angle是求內積(innerproduct).所以我們可以得到:

mathcal{G}^{(1)}(x_{1},x_{1})mathcal{G}^{(1)}(x_{2},x_{2})geqmathcal{G}^{(1)}(x_{1},x_{2})||^2.同理可以推廣到高階.不寫了.

如果是雙波的情況,有mathcal{A}=lambda_{1}E_{1}^{(+)}(x)E_{1}^{(+)}(x)+lambda_{2}E_{2}^{(+)}(x)E_{2}^{(+)}(x),其中xequiv(r,0),xequiv(r,t),所以有mathcal{G}_{11}^{(2)}(0)mathcal{G}_{22}^{(2)}(0)geq|  mathcal{G}_{12}^{(2)}(t)|^{2},按照之前的定義可以知道:mathcal{G}_{ij}^{(2)}(t)=Tr{
ho E_{i}^{(-)}(x)E_{i}^{(-)}(x)E_{j}^{(+)}(x)E_{j}^{(+)}(x)},特別的,mathcal{G}_{ii}^{(2)}是不含時的。如果我們取mathcal{A}=lambda_{1}E_{1}^{(-)}(x)E_{1}^{(+)}(x)+lambda_{2}E_{2}^{(-)}(x)E_{2}^{(+)}(x),這樣不等式得到:

|langle E_{1}^{(-)}(x)E_{1}^{(-)}(x)E_{2}^{(+)}(x)E_{2}^{(-)}(x)|^{2}leqlangle[ E_{1}^{(-)}(x)E_{1}^{(+)}(x)]^{2}
anglelangle[ E_{2}^{(-)}(x)E_{2}^{(+)}(x)]^{2}
angle

  • 一個例子:關聯函數在Youngs雙縫干涉實驗

如圖,電場從兩個縫隙處r_{1},r_{2}入射然後在t疊加到r處,

E^{(+)}(r,t)=E_{1}^{(+)}(r,t)+E_{2}^{(+)}(r,t)

其中E_{i}^{(+)}(r,t)是由i處產生的:

E_{i}^{(+)}(r,t)=E_{i}^{(+)}(r_{i},t-frac{s_{i}}{c})frac{1}{s_{i}}e^{i(k-frac{omega}{c})s_{i}},其中s_{i}=|r_{i}-r|,對於球形波,這裡有零級干涉加強,所以k-frac{omega}{c}=0.

所以上式變成了:E^{(+)}(r,t)=frac{E_{1}^{(+)}(r_{1},t-frac{s_{1}}{c})}{s_{1}}+frac{E_{2}^{(+)}(r_{2},t-frac{s_{2}}{c})}{s_{2}},而縫隙到干涉處的距離其實是差不多的。所以有:s_{1}approx s_{2}approx R,有

E^{(+)}(r,t)=frac{1}{R}[E_{1}^{(+)}(x_{1})+E_{x}^{(+)}(x_{2})],有egin{eqnarray}&& x_{1}=(r_{1},t-frac{s_{1}}{c}) \&& x_{2}=(r_{2},t-frac{s_{2}}{c})end{eqnarray}.我們算是得到了在Youngs雙縫模型里的電場,下面我們用之前的公式求一下:

I=Tr{
ho E^{(-)}(r,t)E^{(+)}(r,t)}

把上面的式子帶進去得到:

I=mathcal{G}^{(+)}(x_{1},x_{1})+mathcal{G}^{(+)}(x_{2},x_{2})+2Re{mathcal{G}^{1}(x_{1},x_{2})},其中R^{-2}因子吸收進歸一化係數里.第一項第二項都是強度項,第三項則是干涉項.在這裡關聯函數寫成mathcal{G}^{(1)}(x_{1},x_{2})=|mathcal{G}^{1}(x_{1},x_{2})|e^{iPsi(x_{1},x_{2})},這樣就可以得到:

I=mathcal{G}^{(+)}(x_{1},x_{1})+mathcal{G}^{(+)}(x_{2},x_{2})+2|mathcal{G}^{(1)}(x_{1},x_{2})|cosPsi(x_{1},x_{2}),實驗中觀察得到的干涉條紋就來自式子里的餘弦項.它就用關聯函數來描述.

  • 一階光場的相干性

光學裡的相干性和之前討論的兩個電場疊加產生干涉條紋有很大的關聯,並且最大光相干性(the highest degree of optical coherence)和條紋的最大可見(fringes with maximum visibility)也有關.如果關聯函數mathcal{G}^{(1)}(x_{1},x_{2})為0那麼也就沒有條紋並且這個場我們稱它不相干(incoherent).所以關聯函數越大那麼相干性越好.但是mathcal{G}^{(1)}(x_{1},x_{2})的尺度也是由之前的不等式限制:|mathcal{G}^{(1)}(x_{1},x_{2})|leq[  mathcal{G}^{(1)}(x_{1},x_{1})mathcal{G}^{(1)}(x_{2},x_{2})]^{1/2},所以最完全相干(full coherence)在這裡:|mathcal{G}^{(1)}(x_{1},x_{2})|=[  mathcal{G}^{(1)}(x_{1},x_{1})mathcal{G}^{(1)}(x_{2},x_{2})]^{1/2},我們給出歸一化關聯函數:g^{(1)}(x_{1},x_{2})=frac{ mathcal{G}^{(1)}(x_{1},x_{2})  }{[mathcal{G}^{(1)}(x_{1},x_{1})mathcal{G}^{(1)}(x_{2},x_{2})]^{1/2}},之前的最完全相干可以表示成|g^{(1)}(x_{1},x_{2})|=1.或者寫成g^{(1)}(x_{1},x_{2})=e^{iPsi(x_{1},x_{2})},所以我們定義襯比度contrast):
u=frac{I_{max}-I_{min}}{I_{max}+I_{min}},聯繫之前可以得到:
u=|frac{    mathcal{G}^{(1)}(x_{1},x_{2})}{(mathcal{G}^{(1)}(x_{1},x_{1})  mathcal{G}^{(1)}(x_{2},x_{2}) })^{1/2}|frac{2(I_{1}I_{2})^{1/2}}{I_{1}+I_{2}}=|g^{(1)}|frac{    2(I_{1}I_{2})^{1/2} }{I_{1}+I_{2}},,這裡我們可以看出一階光學相干性|g^{(1)}|=1和最大襯比度是聯繫在一起的.

讓我們再回到之前的Youngs雙縫干涉實驗,經典理論和量子理論都有解釋。其實量子理論是從光子的概率振幅來考慮的,並且需要考慮到那兩條路徑。所以我們很好奇,如果只有一個光子那麼會怎麼樣?

為了避免去計算衍射,我們假設在縫隙處的光場還是和之前一樣的球形波(Huygensprinciple).我們選取模函數:u_{k}(r)=sqrt{frac{1}{4pi L}}frac{e^{ikcdot r}}{r}hat{e}_{k},L是歸一化半徑,hat{e}_{k}是極化矢量.所以在r,t條件下在屏上的E^{(+)}(r,t)=f(r,t)(a_{1}e^{iks_{1}}+a_{2}e^{iks_{2}}),f(r,t)=i(frac{hbaromega}{2})^{1/2}frac{hat{e}_{k}}{(4pi L)^{1/2}}frac{1}{R}e^{-iomega t},帶入之前我們對強度的公式得到:I(r,t)=eta[Tr{
ho a_{1}^{dagger}a_{1}}+Tr{
ho a_{2}^{dagger}a_{2}}+2|Tr{
ho a_{1}^{dagger}a_{2}}|  cos{Phi}    ],其中

egin{eqnarray}&& Tr{
ho a_{1}^{dagger}a_{2}}=|Tr{
ho a_{1}^{dagger}a_{2}}|e^{iphi} \&& eta=|f(r,t)|^{2} \&&Phi=k(s_{1}-s_{2})+phiend{eqnarray},同樣運用最大幹涉條件知道;k(s_{1}-s_{2})+phi=n2pi.類比之前在量子力學裡面學過的光場有|Psi
angle=f(b^{dagger})|0
angle,其中b^{dagger}a_{1},a_{2}的線性疊加:b^{dagger}=-frac{1}{sqrt{2}}(a_{1}^{dagger}+a_{2}^{dagger}).我們假定通過每個縫隙的光流(我實在不知道怎麼很好的表達intensities,翻譯成強度什麼的都感覺 不好),我們考慮如果只有一個光子的情況那麼有:|1 photon
angle=b^{dagger}|0
angle=frac{1}{sqrt{2}}(|1,0
angle+|0,1
angle).其中本徵態|n_{1},n_{2}
angle表示n_{1}個光子在k_{1}模,n_{2}個光子在k_{2}模.所以光流可以寫成I(r,t)=eta(1+cos{Psi})(frac{1}{4},frac{1}{4},frac{1}{2}).

關於這個的解釋最早是Dirac提出來的,但是其實他的解釋並不很對,所以這裡不提了。

最早對於單光子實驗室1905年由Taylor做的,因為光源強度非常非常低,所以可以看成一個一個光子逐漸的打到屏上.但是可想而是,時間也非常非常長.但是觀測到了干涉條紋.可惜它不能證明這個條紋就是單光子產生的,因為按照統計分布總有一小部分的雙光子入射進去,而這個他並沒有分離開來.後來Grangier用雙光子串聯作為光源,比較好的做到了這一點.

  • 相干場

前面的雙縫例子,入射光源我們假設是理想激光,相干場的波函數可以表達為:|coherent-field
angle=|alpha_{1},alpha_{2}
angle=|alpha_{1}
angle|alpha_{2}
angle,因為這個波函數是內積態,所以可以分成獨立的光路.所以我們有egin{eqnarray}&& |alpha_{1}
angle|alpha_{2}
angle=exp(alpha b^{dagger}-alpha^{*}b)|0
angle \&& =expfrac{1}{sqrt{2}}(alpha a_{1}^{dagger}-alpha^{dagger}a_{2})expfrac{1}{sqrt{2}}(alpha a_{2}^{dagger}-alpha^{dagger}a_{2})|0
angle \&& =|frac{alpha_{1}}{sqrt{2}}
angle|frac{a_{2}}{sqrt{2}}
angleend{eqnarray},光強為I(r,t)=eta(|alpha|^2+|alpha|^2cos{phi})(第一項是強度項,後面是干涉項).我們關注干涉項,它需要我們知道兩個光路的相位關係以及需要儘可能的慢,不然干涉條紋會被破壞掉。

我們考慮到兩個獨立的光源用Fock態來表示:|Psi
angle=|n_{1}
angle|n_{2}
angle,如果這樣表示的話,意味著他們之間沒有關聯,所以也就沒有條紋產生了.

  • 光子測量手段

第一個實驗去測量出光子的關聯性質的是Hanbury-Brown-Twiss實驗.

實驗裝置很簡單.我們假設兩個探測器處的光強分別是:

egin{eqnarray}&& i_{1}=E^{2}sin^{2}(omega t) \&& i_{2}=E^{2}sin^{2}(omega t+phi)=E^{2}(sin(omega t)cos(phi)+sin(phi)cos(omega t))^{2}end{eqnarray},

求一下關聯函數,得到egin{equation}langle i_{1}i_{2}
angle=lim_{T
ightarrowinfty}frac{E^{4}}{T}int_{0}^{T}sin^{2}(omega t)cos(phi)+sin(phi)cos(omega t))^{2}dt=frac{E^{4}}{4}+frac{E^{4}}{8}cos(2phi)end{equation}

一個常量加一個依賴相位的量,但是如果有一個擾動:$Delta i=i-langle i
angle$,我們求一個平均關聯函數:egin{equation}langleDelta i_{1}Delta i_{2}
angle=langle(i_{1}-langle i_{1}
angle)(i_{2}-langle i_{2}
angle)
angle=langle i_{2}i_{2}
angle-langle i_{1}langle i_{2}
angle
angle-langle i_{2}langle i_{1}
angle
angle++langle i_{1}
anglelangle i_{2}
angle=langle i_{1}i_{2}
angle-langle i_{1}
anglelangle i_{2}
angleend{equation},後面兩項分別是frac{E^2}{4},所以結果是langle Delta i_{1}Delta i_{2}
angle=frac{E^4}{8}cos(2phi),

假設在探測光子的時候一個在t被檢測出另一個在t+	au被檢測出.可以寫成光強或光子數關聯函數:egin{eqnarray}&& mathcal{G}^{(2)}(	au)=langle E^{(-)}(t)E^{(-)}(t+	au)E^{(+)}(t)E^{(+)}(t+	au)
angle \&&=langle :I(t)I(t+	au):
angle \&& proptolangle :n(t)n(t+	au):
angleend{eqnarray},其中::表示歸一化指數.我們定義歸一化二階關聯函數:g^{(2)}(	au)=frac{  mathcal{G}^{(2)}(	au)}{| mathcal{G}^{(1)}(0)|^2},很巧的是又有mathcal{G}^{(2)}(	au)=epsilon^{(-)}(t)epsilon^{(-)}(t+	au)epsilon^{(+)}(t+	au)epsilon^{(+)}(t)=[mathcal{G}^{(1)}(0)]^2,所以g^{(2)}(	au)=1,對於擾動的場,我們加入一個概率分布P(epsilon)來描述場E^{(+)}(epsilon,t)epsilon的概率:E^{(+)}(epsilon,t)=-( ifrac{hbaromega}{2epsilon_{0}V})^{1/2}epsilon e^{-iomega t},而對於多模場(multimode -field),二階關聯函數可以寫成:

mathcal{G}^{(2)}(	au)=int P({epsilon_{k}})E^{(-)}(epsilon_{k},t)E^{(-)}(epsilon_{k},t+	au)E^{(+)}(epsilon_{k},t)E^{(+)}(epsilon_{k},t+	au)d^{2}{epsilon_{k}},如果延遲時間	au=0那麼我們可以得到單模場關聯函數(k=1):g^{(0)}(0)=1+frac{int P(epsilon)( |epsilon|^{2}-langle |epsilon|^{2}
angle)^{2}d^2epsilon}{ (langle|epsilon|^2
angle)^2},因為分布概率P(epsilon)>0,所以g^{(2)}(0)geq1.我們算一下上面那個場的分布,直接截圖好了:

因為這個場,所以第一項可以被忽略掉.所以就是mathcal{G}^{(2)}(	au)=mathcal{G}^{(1)}(0)^{2}+mathcal{G}^{(1)}(	au)^2,也可以寫作g^{(2)}(	au)=1+| g^{(1)}(	au)|^2.現在mathcal{G}^{(1)}(	au)是場譜的傅里葉變換:mathcal{S}(omega)=int^{infty}_{-infty}d	au e^{-iomega	au}mathcal{G}^{(1)}(	au),所以如果場是洛倫茲譜(Lorentzianspectrum)可以得到g^{(2)}(	au)=1+e^{-gamma	au},高斯譜(Gaussianspectrum)g^{(2)}(	au)=1+e^{-gamma^{2}	au ^{2}}(這裡不太懂...這是什麼東西...)其中gamma是譜線寬度.

可以從上面 的關聯函數看到,如果	augg	au_{c},那麼關聯函數就g^{(2)}
ightarrow1.在	au<	au_{c}下混浦光(chaotic light)是相干光的兩倍:g^{(2)}(0)_{chaotic}=2g^{{2}}(0)_{coh}.然後給出了光子聚束的定義(直接貼原文)the increased intensity fluctuations in the chaotic light field. There is a high probability that the photon which triggers the counter occurs during a high intensity fluctuation and hence a high probability that a second photon will be detected arbitrarily soon.之後的實驗很好的證明相干光和混浦光的預測.我們注意到之前的分析沒有用到量子化電磁場的知識,而全部都是電磁場模波動振幅的經典分析..

光的二階關聯函數的一些基礎知識就是光子關聯波譜學的基礎(直接翻譯的真是彆扭..).它可以用在測量非常狹窄的線寬情況,g^{2}(	au)可以用電子關聯器和線寬(上面的公式)

  • 壓縮態

我們還是要把之前量子化那一套拿到二階關聯函數里來:回憶一下之前粒子數表示的那一套,我們大概可以猜出它的形式(單模場)g^{(2)}(0)=frac{ langle a^{dagger}a^{dagger}aa
angle}{langle a^{dagger}a
angle}^2=1+frac{ V(n)-ar{n}}{ar{n}^2},其中V(n)=langle(a^{dagger}a)^2
angle-langle a^{dagger}a
angle^2.

我們考慮一個相干態,|alpha
angle,有
ho=|alpha
anglelanglealpha|,g^{(2)}(0)=1,這裡V(n)=ar{n}是光子數的poisson分布.

再考慮一個粒子數態|n
angle,有
ho=|n
anglelangle n|,g^{(2)}(0)=1-frac{1}{n},n>2,還有V(n)=0.

如果有g^{(2)}(	au)<g^{(2)}(0)就會使得粒子儘可能成雙的前進,這就是粒子聚束(photon-bunching),如果反過來就是粒子反聚束(photon-antibunching).實際上,g^{(2)}(	au)<1作為一個標準,那麼若g^{(2)}(0)
ightarrow1展現就是。粒子反聚束.並且這個不太能被經典的解釋,前面有一個公式證明了g^{(2)}(0)geq1,所以如果要g^{(0)}<1那麼必須得場有一個負的概率元(elements of negativeprobability這翻譯..)而這就是粒子反聚束..

我們現在考慮一個壓縮態|alpha,r
angle,我們建立坐標系X_{1}平行於誤差橢圓的短軸,(1)是壓縮方向,(2)是相干激發方向.我們定義2alpha=langle X_{1}
angle +ilangle X_{2}
angle其中角度	heta=	an^{-1}(langle X_{2}
angle/langle X_{1}
angle)

.介紹一下subpoissonian /superpoissonian分布,直接貼wiki

所以光子的壓縮態的方差可以寫出來:frac{V(n)-ar{n}}{ar{n}^2}=frac{|alpha|^2(cosh2r-sinh2rcos2	heta-1)+sinh^2rcosh2r  }{(|alpha|^2+sinh^2r)^2},

如果	heta=frac{pi}{2},可以寫成V(n)=| alpha|^2e^{2r}+2sinh^2rcosh^2r,並且這裡是超泊松分布.

如果	heta=0,寫成V(n)=| alpha|^2e^{-2r}+2sinh^2rcosh^2r,第一項是和最開始泊松分布里粒子數減少有關的,第二項和壓縮真空裡面的額外光子有關..

|alpha|^2geq2sinh^2rcosh^2r,這個是亞泊松分布下的振幅壓縮態,光子數波動的最大減小量可以估計出來.是V(n)approx| alpha|^2e^{-2r}+frac{1}{8}e^{4r},V(n)的最小值在e^{6r}=4| alpha|^2,並且V_{min}(n)approx0.94|alpha|^{4/3}

相空間里的振幅(amplitude)和相(phase)壓縮態表示.a(	heta=0) b.	heta=frac{pi}{2}

下面介紹一下壓縮真空,|alpha
angle=0,有V(n)=ar{n}(1+cosh2r),在這裡壓縮真空展現的是超泊松統計.而粒子數態不表現壓縮性:Delta X_{1}^2=Delta X_{2}^2=2n+1.可以用(X_{1},X_{2})表示成半徑為sqrt{n}寬度為1.

  • 含相關聯函數

偶數階的關聯函數比如二階的mathcal{G}^{(n,n)}(x)是不包含信息的以及可以用來測量光子數的手段.相反奇數階的mathcal{G}^{(n,m)}(x_{1}cdots x_{n},x_{n+1}cdots x_{n+m})是包含信息的(場).

同樣給出一個模型,考慮兩個相同頻率的場E_{1}(r,t),E_{2}(r,t),以及一個束板透射率是eta

下面算起來了,粒子數表象以及是玻色子:egin{eqnarray}&&  E_{1}(r,t)=i(frac{hbaromega}{2Vepsilon_{0}})(ae^{i(kcdot r-omega t)}-a^{dagger}e^{-i(kcdot-omega t)}) \ &&E_{2}(r,t)=i(frac{hbaromega}{2Vepsilon_{0}})(be^{i(kcdot r-omega t)}-b^{dagger}e^{-i(kcdot-omega t)})end{eqnarray}加起來

egin{eqnarray}&& E_{1}(r,t)=i(frac{hbaromega}{2Vepsilon_{0}})(ae^{i(kcdot r-omega t)}-a^{dagger}e^{-i(kcdot-omega t)}) \&& c=sqrt{eta a}+isqrt{1-eta}bend{eqnarray},

我們也可以得到

langle c^{dagger}c
angle=etalangle a^{dagger}a
angle+(1-eta)langle b^{dagger}b
angle-isqrt{eta(1-eta)}(langle a 
anglelangle b^{dagger}
angle-langle a^{dagger}
anglelangle b
angle),

我們假定E_{2}局域場假定它是有相干態,所以第一項可以忽略掉:

langle c^{dagger}c
angleapprox(1-eta)|eta|^2+|eta|sqrt{eta(1-eta)}langle X_{	heta+pi/2}
angle,X_{	heta}=ae^{-i	heta}+a^{dagger}e^{i	heta}.	hetaeta的相.

再看光流的微擾問題,後面的平方項是有c^{dagger}c確定的.V(n_{c})approx(1-eta)^2|eta|^2+|eta|^2eta(1-eta)V(X_{	heta+pi/2})

第一項表示了反射局域諧振子強度的微擾,如果這一項被減少了,光流的擾動由V(X_{	heta+pi/2})決定的.

  • 光子計數

考慮一個計數器,在dt里計數的概率是Delta p(t)=alpha I(t)dt,alpha是測量靈敏度並且依賴於入射光的範圍,假定最開始沒有微擾的I(t),現在1-Delta p(t)表示在dt範圍里沒有發生計數,在t
ightarrow t+T內的非計數概率為:

prod_{t}^{t+T}[1-Delta p(t)]approxprod_{t}^{t+T}exp[-Delta p(t)]=exp[-sum_{t}^{t+T}Delta p(t)]exp[-int_{t}^{t+T}dp(t)],所以P_{0}(T+t,t)=exp[-alphadisplaystyle int_{t}^{t+T}I(t)dt],所以有一個粒子計數在t+T就是P_{1}(T+t,t)=[alphaint_{t}^{t+T}dp(t^{})]exp[-int_{t}^{t+T}dp(t)],推廣一下n個計數就是:

P_{n}(t,t+T)frac{1}{n!}[alpha Tar{I}(t,T)^{n}]exp[-alpha Tar{I}(t,T)],(ar{I}(t,T)=frac{1}{T}int_{t}^{t+T}I(t)dt是平均強度),因為ar{I}(t,T)是變化的按照開始計數的時間,所以P_{n}(T)是把所有開始的時間的值加起來的平均:

egin{eqnarray}&& P_{n}(T)=langle P_{n}(t,T)
angle \&& =langlefrac{[alphaar{I}(t,T)T]^n}{n!}exp[-alphaar{I}(t,T)]T
angleend{eqnarray}

我們注意到一個廣義的分布函數:Q(lambda,T)=sum_{n=0}^{infty}(1-lambda)^n P_{n}(T),可以得到sum_{n=0}^{infty}n(n-1)cdots(n-k)P_{n}(T)=(-1)^kfrac{partial^k}{partiallambda^k}Q(lambda,T)|_{lambda=0}

  • 恆定強度

如果這個量I(t,T)=I是獨立於t,T,所以應該這麼寫:P_{n}(T)=frac{ar{n}^n}{n!}exp(-ar{n}),其中ar{n}=alpha IT,這個泊松分布里方差是V(n)=ar{n}

  • 微擾強度——短時近似

強度開始微擾時,因為計數時間T要小於相干時間	au_{c},所以在T內,強度是不變的:I(t,T)=I(t),我們可以得到:P_{n}(T)=int_{0}^{infty}frac{[alpha I(t)T]^n}{n!}e^{-alpha I(t)T}p(I(t))dI(t),其中p(I(t))是分布.

下面我們用隨機變數I代替ar{I}(t),平均粒子數:ar{n}=sum_{n=0}^{infty}nP_{n}(T)=int_{0}^{infty}alpha TIp(I)dI=alphalangle I
angle,其中langle I^m
angle=int_{0}^{infty}I^n p(I)dI,

平方粒子數為ar{n^2}=sum_{n=0}^{infty}n^2P_{n}(T)=int_{0}^{infty}(alpha^2 T^2I^2+alpha TI)p(I)dI=alpha^2 T^2langle I^2
angle+alpha Tlangle I
angle,所以方差V(n)=ar{n^2}-ar{n}^2=alpha Tlangle I
angle+alpha^2T^2(langle I^2
angle-langle I
angle^2).

我們注意到V(n)>ar{n}除非p(I)Dirac函數delta(I-I_{0}),但是我們在量子層次會看到V(n)<ar{n}.

熱光源場有如下的一些性質:

p(I)=frac{1}{I_0}exp(frac{-I}{I_0}),以及由langle I 
angle=n!I_{0}^{n},所以平均值和方差分別是:

egin{eqnarray}&& ar{n}=alpha T_{0} \&& V(n)=ar{n}(1+ar{n})end{eqnarray}

可以得到光子數分布分別是:

egin{eqnarray}&& P_{n}(T)=frac{(alpha T)^n}{I_0n!}int_{0}^{infty}I^nexp[-I(alpha T +frac{1}{I_0})]dl \&& =frac{1}{(1+ar{n})}(frac{ar{n}}{1+ar{n}})^nend{eqnarray},

假如I(t,T)按照I(t)=I_{0}e^{-lambda t}損耗,那麼在一段時間內積分得到I(t,T)=frac{I_{0}}{T}int_{t}^{t+T}e^{-lambda t}dt=frac{I(t)}{lambda T}(1-e^{-lambda T}),如果我們取短時近似,發現和上面提到的一樣是恆定的.

  • 光子計數(量子)類比經典的情況:P_{n}(T)=langle:frac{[alpha ar{I}(T)T]^n}{n!}exp[-alpha I(T)T]:
angle,其中有I(T)=frac{1}{T}int_{0}^{T}I(t)dt=frac{1}{T}int_{0}^{T}E^{(-)}(r,t)E^{(+)}(r,t)dt::表示這個算符的指數.我們可以寫出它的量子形態:P_{n}(T)=Tr(
ho:frac{[mu(T)a^{dagger}a]^n}{n!}exp[-a^{dagger}amu(T)]:),mu代表在單光子場下在時間T內測量出一個光子的概率,並且是取決於系統狀態的,如果是開系就是mu(T)=lambda T,閉系就是mu(T)=(1-e^{-lambda T})

光子計數分布和
ho對角矩陣元P_{n}=langle n|
ho|n
angleP_{m}(T)=sum_{n}P_{n}frac{[mu(T)]^m}{m!}langle n|sum_{l=0}^{infty}frac{mu(T)^l}{l!}a^{dagger m+1}a^{m+1}|n
angle=sum_{n=m}^{infty}P_{n}sum_{l=0}^{n-m}(-1)^lfrac{mu(T)^l}{l!}frac{n!}{(n-m-l)!}=egin{equation}sum_{n=m}^{infty}P_{n}egin{pmatrix} n\ m end{pmatrix}[mu(T)]^m[1-mu(T)]^{n-m}end{equation},其中egin{pmatrix} n\ m end{pmatrix}=frac{n!}{m!(n-m)!},這也就是伯努利分布.

相干光:egin{eqnarray}&& P_{n}=frac{ar{n}^n}{n!}exp(-ar{n}) \&& P_{m}(T)=frac{[mu(Tar{n})]^m}{m!}exp(-mu(T)ar{n})end{eqnarray}

混浦光:

egin{eqnarray}&& P_{n}=frac{(ar{n})^n}{(1+ar{n})^{1+n}} \&& P_{m}(T)=frac{[mu(T)ar{n}]^m}{[1+mu(T)ar{n}]^{1+m}}end{eqnarray}

前面的知識告訴我們平均粒子數ar{n}和平均計數粒子ar{m}的關係ar{m}=mu(T)ar{n}

對於單模場sum m(m-1)frac{P_{m}(T)}{ar{m}^2}=sum_{m}(n-1)frac{P_n}{ar{n}^2}=g^{(2)}(0),這也是一種求關聯函數的方法.

看到兩個講壓縮光很好的東西,貼在這裡.

Squeezed States of Light

Definition: nonclassical states of light with noise below the standard quantum limit in one quadrature component

Squeezed states of light (or squeezed light) are a kind of nonclassical light and constitute an interesting subject of quantum optics, the experimental investigation of which began in the 1980s.

Squeezed light is best understood by considering complex phasors for the representation of the state of light in one mode of the optical field. Classically, such a state can be represented by a certain phasor (or its end point in the complex plane). According to quantum optics, however, there is a quantum uncertainty, and any measurement of the complex amplitude of the light field can deliver different values within an uncertainty region. Moreover, there is an uncertainty relation for the quadrature components of the light field, saying that the product of the uncertainties in both components is at least some quantity times Plancks constant.

Glaubers coherent states have circularly symmetric uncertainty regions, so that the uncertainty relation dictates some minimum noise amplitudes e.g. for the amplitude and phase. A further reduction in, e.g., amplitude noise is possible only by 「squeezing」 the uncertainty region, reducing its width in the amplitude direction while increasing it in the orthogonal direction, so that the phase uncertainty is increased. Such light is called amplitude-squeezed (see Figure 1, left). Conversely, phase-squeezed light (Figure 1, middle) has decreased phase fluctuations at the expense of increased amplitude fluctuations.

Of course, there are also squeezed states where the orientation of the uncertainty region is different from the cases shown, or where the shape of the uncertainty region is different from that of an ellipse. In any case, some noise component is below the standard quantum limit.

There is also the so-called squeezed vacuum (Figure 1, right), where the center of the uncertainty region (corresponding to the average amplitude) is at the origin of the coordinate system, and the fluctuations are reduced in some direction. The mean photon number is larger than zero in this case; a squeezed vacuum is a 「vacuum」 only in the sense that the average amplitude (but not the average photon number) is zero. Squeezed light with a non-zero average amplitude is also called bright squeezed light.

Quantum noise also leads to fluctuations of the polarization, which are reduced in polarization-squeezed light.

? Generation of Squeezed Light

Squeezed light can be generated from light in a coherent state or vacuum state by using certain optical nonlinear interactions. For example, an optical parametric amplifier with a vacuum input can generate a squeezed vacuum with a reduction in the noise of one quadrature components by the order of 10 dB. A lower degree of squeezing in bright amplitude-squeezed light can under some circumstances be obtained with frequency doubling. The Kerr nonlinearity in optical fibers also allows the generation of amplitude-squeezed light. Semiconductor lasers can generate amplitude-squeezed light when operated with a carefully stabilized pump current. Squeezing can also arise from atom-light interactions.

? Applications

In principle, squeezed light can be used in a number of areas, as it allows for measurements with reduced quantum noise. An example is the ultraprecise measurement of lengths for the detection of gravitational waves with large-scale interferometers. However, the use of squeezed light is not very widespread, basically because it is plagued with various difficulties. For example, any optical losses bring a squeezed state of light closer to a coherent state, i.e. tend to destroy the nonclassical properties. At least in fundamental quantum optics research, however, squeezed states of light play an important role.

Nonclassical light is light with nonclassical quantum noise properties, which can be understood only on the basis of quantum optics. The most common forms of nonclassical light are the following:

  • Squeezed light exhibits reduced noise in one quadrature component. The most familiar kinds of squeezed light have either reduced intensity noise or reduced phase noise, with increased noise of the other kind.
  • Fock states (also called photon number states) have a well-defined number of photons (stored e.g. in a cavity), whereas the phase is totally undefined. A special case is that of a single-photon state, as generated on demand by a photon gun.
  • Nonclassical light sometimes exhibits photon antibunching, i.e. a reduced probability of two photons being detected within a short time interval, or sub-Poissonian photon statistics.
  • Also, there are states with special correlations between two or more different light beams. For example, signal and idler in parametric amplification exhibit strong intensity noise correlations, because signal/idler photons are generated in pairs. Nonclassical correlations of different light beams also occur in the context of quantum nondemolition measurements.

Nonclassical light is often generated either in nonlinear devices such as in sub-threshold optical parametric oscillators or frequency doublers, or in systems with only a single atom or ion (or just a few such emitters), such as a single-atom laser. It is relevant in fundamental quantum physics and also in the context of some high-precision measurements, such as for gravitational wave detection.

? Bibliography

[1]R. J. Glauber, 「Coherent and incoherent states of the radiation field」, Phys. Rev. 131 (6), 2766 (1963)[2]D. Walls, 「Squeezed states of light」, Nature 306, 141 (1983)[3]H. J. Kimble and D. Walls (eds.), Special Issue on Squeezed Light, J. Opt. Soc. Am B 4 (10) (1987)

references &&further reading

1. R.J. Glauber: Phys. Rev. 130, 2529 (1963)2. M. Born, E. Wolf: Principles of Optics, 3rd edn. (Pergamon, London 1965)3. P.A.M. Dirac: The Principles of Quantum Mechanics, 3rd edn. (Clarendon, London 1947)4. G.I. Taylor: Proc. Cam. Phil. Soc. Math. Phys. Sci. 15, 114 (1909)5. P. Grangier, G. Roger, A. Aspect: Europhys. Lett. 1, 173 (1986)6. R.F. Pfleegor, L. Mandel: Phys. Rev. 159, 1084 (1967)7. R. Hanbury-Brown, R.W. Twiss: Nature 177, 27 (1956)8. F.T. Arecchi, E. Gatti, A. Sona: Phys. Rev. Lett. 20, 27 (1966); F.T. Arecchi, Phys. Lett. 16,32 (1966)9. W. Demtroeder: Laser Spectroscopy, 2nd edn. (Springer, Berlin, Heidelberg 1993)10. L. Mandel: Phys. Rev. Lett. 49, 136 (1982)11. H.P. Yuen, J.H. Shapiro: IEEE Trans. IT 26, 78 (1980)12. L. Mandel: Progress in Optics 2, 183 (North-Holland, Amsterdam 1963)13. E.R. Pike: In Quantum Optics, ed. by R.J. Glauber (Academic, New York 1970)14. P.L. Kelley, W.H. Kleiner: Phys. Rev. 136, 316 (1964)15. M.O. Scully, W.E. Lamb: Phys. Rev. 179, 368 (1969)16. H.J. Carmichael: J. Opt. Soc. Am. B 4, 1588 (1987)17. D. Mettzer, L. Mandel: IEEE J. QE-6, 661 (1970)Further ReadingMeystre, P., M. Sargent III: Elements of Quantum Optics, 2nd edn. (Springer, Berlin,Heidelberg 1991)Perina, J.: Coherence of Light (van Nostrand Reinhold, London 1972)

推薦閱讀:

相機里的光圈和快門長什麼樣?
為什麼是三原色,而不是四原色、五原色?
如何通俗地解釋「壓縮光(Squeezed light)」?它在傳統通訊、量子通訊和加密技術中有何應用?
普通激光手電筒如何加大功率使其具有破壞力?
在設計脈衝激光器時需要注意什麼問題?

TAG:物理学 | 量子信息 | 光学 |